diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5347 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9997351694915254,
+  "eval_steps": 200,
+  "global_step": 755,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0013241525423728813,
+      "grad_norm": 26.125,
+      "learning_rate": 2.6315789473684213e-07,
+      "loss": 3.7676,
+      "step": 1
+    },
+    {
+      "epoch": 0.0013241525423728813,
+      "eval_AskDoctor-Chinese_loss": 3.761261463165283,
+      "eval_AskDoctor-Chinese_runtime": 15.9737,
+      "eval_AskDoctor-Chinese_samples_per_second": 1.878,
+      "eval_AskDoctor-Chinese_steps_per_second": 1.878,
+      "step": 1
+    },
+    {
+      "epoch": 0.0026483050847457626,
+      "grad_norm": 27.375,
+      "learning_rate": 5.263157894736843e-07,
+      "loss": 3.8885,
+      "step": 2
+    },
+    {
+      "epoch": 0.003972457627118644,
+      "grad_norm": 28.125,
+      "learning_rate": 7.894736842105263e-07,
+      "loss": 3.8601,
+      "step": 3
+    },
+    {
+      "epoch": 0.005296610169491525,
+      "grad_norm": 26.25,
+      "learning_rate": 1.0526315789473685e-06,
+      "loss": 3.8192,
+      "step": 4
+    },
+    {
+      "epoch": 0.0066207627118644065,
+      "grad_norm": 29.0,
+      "learning_rate": 1.3157894736842106e-06,
+      "loss": 3.7572,
+      "step": 5
+    },
+    {
+      "epoch": 0.007944915254237288,
+      "grad_norm": 30.5,
+      "learning_rate": 1.5789473684210526e-06,
+      "loss": 3.9407,
+      "step": 6
+    },
+    {
+      "epoch": 0.009269067796610169,
+      "grad_norm": 30.25,
+      "learning_rate": 1.8421052631578948e-06,
+      "loss": 3.9034,
+      "step": 7
+    },
+    {
+      "epoch": 0.01059322033898305,
+      "grad_norm": 30.5,
+      "learning_rate": 2.105263157894737e-06,
+      "loss": 3.7951,
+      "step": 8
+    },
+    {
+      "epoch": 0.011917372881355932,
+      "grad_norm": 26.0,
+      "learning_rate": 2.368421052631579e-06,
+      "loss": 3.7,
+      "step": 9
+    },
+    {
+      "epoch": 0.013241525423728813,
+      "grad_norm": 28.625,
+      "learning_rate": 2.631578947368421e-06,
+      "loss": 3.6931,
+      "step": 10
+    },
+    {
+      "epoch": 0.014565677966101694,
+      "grad_norm": 25.125,
+      "learning_rate": 2.8947368421052634e-06,
+      "loss": 3.5419,
+      "step": 11
+    },
+    {
+      "epoch": 0.015889830508474576,
+      "grad_norm": 25.5,
+      "learning_rate": 3.157894736842105e-06,
+      "loss": 3.6584,
+      "step": 12
+    },
+    {
+      "epoch": 0.017213983050847457,
+      "grad_norm": 23.875,
+      "learning_rate": 3.421052631578948e-06,
+      "loss": 3.5614,
+      "step": 13
+    },
+    {
+      "epoch": 0.018538135593220338,
+      "grad_norm": 23.875,
+      "learning_rate": 3.6842105263157896e-06,
+      "loss": 3.5783,
+      "step": 14
+    },
+    {
+      "epoch": 0.01986228813559322,
+      "grad_norm": 23.25,
+      "learning_rate": 3.947368421052632e-06,
+      "loss": 3.5407,
+      "step": 15
+    },
+    {
+      "epoch": 0.0211864406779661,
+      "grad_norm": 22.0,
+      "learning_rate": 4.210526315789474e-06,
+      "loss": 3.3536,
+      "step": 16
+    },
+    {
+      "epoch": 0.022510593220338982,
+      "grad_norm": 20.0,
+      "learning_rate": 4.473684210526316e-06,
+      "loss": 3.2685,
+      "step": 17
+    },
+    {
+      "epoch": 0.023834745762711863,
+      "grad_norm": 19.5,
+      "learning_rate": 4.736842105263158e-06,
+      "loss": 3.3136,
+      "step": 18
+    },
+    {
+      "epoch": 0.025158898305084745,
+      "grad_norm": 18.875,
+      "learning_rate": 5e-06,
+      "loss": 3.3371,
+      "step": 19
+    },
+    {
+      "epoch": 0.026483050847457626,
+      "grad_norm": 26.125,
+      "learning_rate": 5.263157894736842e-06,
+      "loss": 3.2935,
+      "step": 20
+    },
+    {
+      "epoch": 0.027807203389830507,
+      "grad_norm": 18.375,
+      "learning_rate": 5.526315789473685e-06,
+      "loss": 3.1542,
+      "step": 21
+    },
+    {
+      "epoch": 0.02913135593220339,
+      "grad_norm": 16.125,
+      "learning_rate": 5.789473684210527e-06,
+      "loss": 3.0697,
+      "step": 22
+    },
+    {
+      "epoch": 0.03045550847457627,
+      "grad_norm": 17.75,
+      "learning_rate": 6.0526315789473685e-06,
+      "loss": 3.0135,
+      "step": 23
+    },
+    {
+      "epoch": 0.03177966101694915,
+      "grad_norm": 16.5,
+      "learning_rate": 6.31578947368421e-06,
+      "loss": 2.9917,
+      "step": 24
+    },
+    {
+      "epoch": 0.033103813559322036,
+      "grad_norm": 14.5625,
+      "learning_rate": 6.578947368421054e-06,
+      "loss": 2.9261,
+      "step": 25
+    },
+    {
+      "epoch": 0.034427966101694914,
+      "grad_norm": 20.375,
+      "learning_rate": 6.842105263157896e-06,
+      "loss": 2.8072,
+      "step": 26
+    },
+    {
+      "epoch": 0.0357521186440678,
+      "grad_norm": 13.375,
+      "learning_rate": 7.1052631578947375e-06,
+      "loss": 2.876,
+      "step": 27
+    },
+    {
+      "epoch": 0.037076271186440676,
+      "grad_norm": 12.5,
+      "learning_rate": 7.368421052631579e-06,
+      "loss": 2.7867,
+      "step": 28
+    },
+    {
+      "epoch": 0.03840042372881356,
+      "grad_norm": 14.625,
+      "learning_rate": 7.631578947368423e-06,
+      "loss": 2.6454,
+      "step": 29
+    },
+    {
+      "epoch": 0.03972457627118644,
+      "grad_norm": 16.625,
+      "learning_rate": 7.894736842105265e-06,
+      "loss": 2.6395,
+      "step": 30
+    },
+    {
+      "epoch": 0.041048728813559324,
+      "grad_norm": 10.1875,
+      "learning_rate": 8.157894736842106e-06,
+      "loss": 2.5909,
+      "step": 31
+    },
+    {
+      "epoch": 0.0423728813559322,
+      "grad_norm": 12.4375,
+      "learning_rate": 8.421052631578948e-06,
+      "loss": 2.5511,
+      "step": 32
+    },
+    {
+      "epoch": 0.043697033898305086,
+      "grad_norm": 10.375,
+      "learning_rate": 8.68421052631579e-06,
+      "loss": 2.4177,
+      "step": 33
+    },
+    {
+      "epoch": 0.045021186440677964,
+      "grad_norm": 8.5625,
+      "learning_rate": 8.947368421052632e-06,
+      "loss": 2.4405,
+      "step": 34
+    },
+    {
+      "epoch": 0.04634533898305085,
+      "grad_norm": 8.5625,
+      "learning_rate": 9.210526315789474e-06,
+      "loss": 2.5073,
+      "step": 35
+    },
+    {
+      "epoch": 0.04766949152542373,
+      "grad_norm": 7.625,
+      "learning_rate": 9.473684210526315e-06,
+      "loss": 2.4031,
+      "step": 36
+    },
+    {
+      "epoch": 0.04899364406779661,
+      "grad_norm": 9.375,
+      "learning_rate": 9.736842105263159e-06,
+      "loss": 2.4698,
+      "step": 37
+    },
+    {
+      "epoch": 0.05031779661016949,
+      "grad_norm": 6.75,
+      "learning_rate": 1e-05,
+      "loss": 2.4379,
+      "step": 38
+    },
+    {
+      "epoch": 0.051641949152542374,
+      "grad_norm": 7.59375,
+      "learning_rate": 9.999952004474853e-06,
+      "loss": 2.3739,
+      "step": 39
+    },
+    {
+      "epoch": 0.05296610169491525,
+      "grad_norm": 6.375,
+      "learning_rate": 9.999808018820836e-06,
+      "loss": 2.433,
+      "step": 40
+    },
+    {
+      "epoch": 0.05429025423728814,
+      "grad_norm": 5.96875,
+      "learning_rate": 9.999568045802216e-06,
+      "loss": 2.4569,
+      "step": 41
+    },
+    {
+      "epoch": 0.055614406779661014,
+      "grad_norm": 5.59375,
+      "learning_rate": 9.99923209002605e-06,
+      "loss": 2.3297,
+      "step": 42
+    },
+    {
+      "epoch": 0.0569385593220339,
+      "grad_norm": 5.5,
+      "learning_rate": 9.998800157942083e-06,
+      "loss": 2.3064,
+      "step": 43
+    },
+    {
+      "epoch": 0.05826271186440678,
+      "grad_norm": 5.375,
+      "learning_rate": 9.99827225784264e-06,
+      "loss": 2.2746,
+      "step": 44
+    },
+    {
+      "epoch": 0.05958686440677966,
+      "grad_norm": 4.9375,
+      "learning_rate": 9.997648399862457e-06,
+      "loss": 2.2981,
+      "step": 45
+    },
+    {
+      "epoch": 0.06091101694915254,
+      "grad_norm": 5.5625,
+      "learning_rate": 9.99692859597849e-06,
+      "loss": 2.0856,
+      "step": 46
+    },
+    {
+      "epoch": 0.062235169491525424,
+      "grad_norm": 5.71875,
+      "learning_rate": 9.996112860009689e-06,
+      "loss": 2.1083,
+      "step": 47
+    },
+    {
+      "epoch": 0.0635593220338983,
+      "grad_norm": 5.53125,
+      "learning_rate": 9.995201207616718e-06,
+      "loss": 2.1841,
+      "step": 48
+    },
+    {
+      "epoch": 0.06488347457627118,
+      "grad_norm": 4.65625,
+      "learning_rate": 9.994193656301676e-06,
+      "loss": 2.3351,
+      "step": 49
+    },
+    {
+      "epoch": 0.06620762711864407,
+      "grad_norm": 5.1875,
+      "learning_rate": 9.993090225407743e-06,
+      "loss": 2.2625,
+      "step": 50
+    },
+    {
+      "epoch": 0.06753177966101695,
+      "grad_norm": 13.0625,
+      "learning_rate": 9.991890936118817e-06,
+      "loss": 2.1591,
+      "step": 51
+    },
+    {
+      "epoch": 0.06885593220338983,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.990595811459109e-06,
+      "loss": 2.2935,
+      "step": 52
+    },
+    {
+      "epoch": 0.0701800847457627,
+      "grad_norm": 4.46875,
+      "learning_rate": 9.98920487629269e-06,
+      "loss": 2.4604,
+      "step": 53
+    },
+    {
+      "epoch": 0.0715042372881356,
+      "grad_norm": 4.5625,
+      "learning_rate": 9.987718157323026e-06,
+      "loss": 2.249,
+      "step": 54
+    },
+    {
+      "epoch": 0.07282838983050847,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.986135683092461e-06,
+      "loss": 2.2686,
+      "step": 55
+    },
+    {
+      "epoch": 0.07415254237288135,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.98445748398167e-06,
+      "loss": 2.291,
+      "step": 56
+    },
+    {
+      "epoch": 0.07547669491525423,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.982683592209069e-06,
+      "loss": 2.1275,
+      "step": 57
+    },
+    {
+      "epoch": 0.07680084745762712,
+      "grad_norm": 12.875,
+      "learning_rate": 9.980814041830203e-06,
+      "loss": 2.1489,
+      "step": 58
+    },
+    {
+      "epoch": 0.078125,
+      "grad_norm": 4.375,
+      "learning_rate": 9.978848868737099e-06,
+      "loss": 2.271,
+      "step": 59
+    },
+    {
+      "epoch": 0.07944915254237288,
+      "grad_norm": 4.84375,
+      "learning_rate": 9.976788110657558e-06,
+      "loss": 2.3377,
+      "step": 60
+    },
+    {
+      "epoch": 0.08077330508474577,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.974631807154447e-06,
+      "loss": 2.3359,
+      "step": 61
+    },
+    {
+      "epoch": 0.08209745762711865,
+      "grad_norm": 4.25,
+      "learning_rate": 9.972379999624935e-06,
+      "loss": 2.2142,
+      "step": 62
+    },
+    {
+      "epoch": 0.08342161016949153,
+      "grad_norm": 12.375,
+      "learning_rate": 9.970032731299697e-06,
+      "loss": 2.1888,
+      "step": 63
+    },
+    {
+      "epoch": 0.0847457627118644,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.967590047242082e-06,
+      "loss": 2.3571,
+      "step": 64
+    },
+    {
+      "epoch": 0.0860699152542373,
+      "grad_norm": 4.25,
+      "learning_rate": 9.96505199434725e-06,
+      "loss": 2.1241,
+      "step": 65
+    },
+    {
+      "epoch": 0.08739406779661017,
+      "grad_norm": 4.65625,
+      "learning_rate": 9.962418621341275e-06,
+      "loss": 2.0733,
+      "step": 66
+    },
+    {
+      "epoch": 0.08871822033898305,
+      "grad_norm": 9.75,
+      "learning_rate": 9.959689978780207e-06,
+      "loss": 2.1612,
+      "step": 67
+    },
+    {
+      "epoch": 0.09004237288135593,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.956866119049095e-06,
+      "loss": 2.1519,
+      "step": 68
+    },
+    {
+      "epoch": 0.09136652542372882,
+      "grad_norm": 9.125,
+      "learning_rate": 9.953947096360996e-06,
+      "loss": 2.1352,
+      "step": 69
+    },
+    {
+      "epoch": 0.0926906779661017,
+      "grad_norm": 4.25,
+      "learning_rate": 9.950932966755917e-06,
+      "loss": 2.1954,
+      "step": 70
+    },
+    {
+      "epoch": 0.09401483050847458,
+      "grad_norm": 4.03125,
+      "learning_rate": 9.947823788099754e-06,
+      "loss": 2.1761,
+      "step": 71
+    },
+    {
+      "epoch": 0.09533898305084745,
+      "grad_norm": 3.921875,
+      "learning_rate": 9.94461962008317e-06,
+      "loss": 2.1397,
+      "step": 72
+    },
+    {
+      "epoch": 0.09666313559322035,
+      "grad_norm": 4.625,
+      "learning_rate": 9.941320524220455e-06,
+      "loss": 2.1281,
+      "step": 73
+    },
+    {
+      "epoch": 0.09798728813559322,
+      "grad_norm": 7.96875,
+      "learning_rate": 9.937926563848345e-06,
+      "loss": 1.8948,
+      "step": 74
+    },
+    {
+      "epoch": 0.0993114406779661,
+      "grad_norm": 4.75,
+      "learning_rate": 9.934437804124807e-06,
+      "loss": 2.3239,
+      "step": 75
+    },
+    {
+      "epoch": 0.10063559322033898,
+      "grad_norm": 3.984375,
+      "learning_rate": 9.93085431202778e-06,
+      "loss": 2.3193,
+      "step": 76
+    },
+    {
+      "epoch": 0.10195974576271187,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.9271761563539e-06,
+      "loss": 2.0956,
+      "step": 77
+    },
+    {
+      "epoch": 0.10328389830508475,
+      "grad_norm": 4.65625,
+      "learning_rate": 9.92340340771717e-06,
+      "loss": 2.3226,
+      "step": 78
+    },
+    {
+      "epoch": 0.10460805084745763,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.919536138547611e-06,
+      "loss": 2.1505,
+      "step": 79
+    },
+    {
+      "epoch": 0.1059322033898305,
+      "grad_norm": 6.1875,
+      "learning_rate": 9.915574423089872e-06,
+      "loss": 2.0939,
+      "step": 80
+    },
+    {
+      "epoch": 0.1072563559322034,
+      "grad_norm": 12.75,
+      "learning_rate": 9.911518337401792e-06,
+      "loss": 2.2263,
+      "step": 81
+    },
+    {
+      "epoch": 0.10858050847457627,
+      "grad_norm": 4.03125,
+      "learning_rate": 9.907367959352964e-06,
+      "loss": 2.1133,
+      "step": 82
+    },
+    {
+      "epoch": 0.10990466101694915,
+      "grad_norm": 3.984375,
+      "learning_rate": 9.903123368623216e-06,
+      "loss": 2.1183,
+      "step": 83
+    },
+    {
+      "epoch": 0.11122881355932203,
+      "grad_norm": 4.46875,
+      "learning_rate": 9.898784646701087e-06,
+      "loss": 2.1461,
+      "step": 84
+    },
+    {
+      "epoch": 0.11255296610169492,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.894351876882277e-06,
+      "loss": 2.1963,
+      "step": 85
+    },
+    {
+      "epoch": 0.1138771186440678,
+      "grad_norm": 5.09375,
+      "learning_rate": 9.889825144268029e-06,
+      "loss": 2.1679,
+      "step": 86
+    },
+    {
+      "epoch": 0.11520127118644068,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.88520453576351e-06,
+      "loss": 2.1925,
+      "step": 87
+    },
+    {
+      "epoch": 0.11652542372881355,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.88049014007613e-06,
+      "loss": 2.1166,
+      "step": 88
+    },
+    {
+      "epoch": 0.11784957627118645,
+      "grad_norm": 3.796875,
+      "learning_rate": 9.875682047713847e-06,
+      "loss": 2.0244,
+      "step": 89
+    },
+    {
+      "epoch": 0.11917372881355932,
+      "grad_norm": 4.125,
+      "learning_rate": 9.87078035098343e-06,
+      "loss": 2.1427,
+      "step": 90
+    },
+    {
+      "epoch": 0.1204978813559322,
+      "grad_norm": 4.25,
+      "learning_rate": 9.865785143988684e-06,
+      "loss": 2.2495,
+      "step": 91
+    },
+    {
+      "epoch": 0.12182203389830508,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.860696522628638e-06,
+      "loss": 2.1213,
+      "step": 92
+    },
+    {
+      "epoch": 0.12314618644067797,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.855514584595719e-06,
+      "loss": 2.1922,
+      "step": 93
+    },
+    {
+      "epoch": 0.12447033898305085,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.850239429373855e-06,
+      "loss": 2.16,
+      "step": 94
+    },
+    {
+      "epoch": 0.12579449152542374,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.84487115823659e-06,
+      "loss": 2.1521,
+      "step": 95
+    },
+    {
+      "epoch": 0.1271186440677966,
+      "grad_norm": 4.375,
+      "learning_rate": 9.839409874245118e-06,
+      "loss": 2.1293,
+      "step": 96
+    },
+    {
+      "epoch": 0.1284427966101695,
+      "grad_norm": 8.6875,
+      "learning_rate": 9.833855682246319e-06,
+      "loss": 2.1652,
+      "step": 97
+    },
+    {
+      "epoch": 0.12976694915254236,
+      "grad_norm": 4.03125,
+      "learning_rate": 9.828208688870736e-06,
+      "loss": 2.0963,
+      "step": 98
+    },
+    {
+      "epoch": 0.13109110169491525,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.822469002530531e-06,
+      "loss": 2.1573,
+      "step": 99
+    },
+    {
+      "epoch": 0.13241525423728814,
+      "grad_norm": 3.75,
+      "learning_rate": 9.816636733417413e-06,
+      "loss": 2.1625,
+      "step": 100
+    },
+    {
+      "epoch": 0.133739406779661,
+      "grad_norm": 4.25,
+      "learning_rate": 9.810711993500506e-06,
+      "loss": 2.1151,
+      "step": 101
+    },
+    {
+      "epoch": 0.1350635593220339,
+      "grad_norm": 4.125,
+      "learning_rate": 9.804694896524215e-06,
+      "loss": 2.0776,
+      "step": 102
+    },
+    {
+      "epoch": 0.1363877118644068,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.79858555800603e-06,
+      "loss": 2.0534,
+      "step": 103
+    },
+    {
+      "epoch": 0.13771186440677965,
+      "grad_norm": 3.953125,
+      "learning_rate": 9.792384095234312e-06,
+      "loss": 1.994,
+      "step": 104
+    },
+    {
+      "epoch": 0.13903601694915255,
+      "grad_norm": 4.5,
+      "learning_rate": 9.78609062726605e-06,
+      "loss": 1.9252,
+      "step": 105
+    },
+    {
+      "epoch": 0.1403601694915254,
+      "grad_norm": 3.90625,
+      "learning_rate": 9.779705274924563e-06,
+      "loss": 2.1531,
+      "step": 106
+    },
+    {
+      "epoch": 0.1416843220338983,
+      "grad_norm": 3.828125,
+      "learning_rate": 9.773228160797187e-06,
+      "loss": 2.1426,
+      "step": 107
+    },
+    {
+      "epoch": 0.1430084745762712,
+      "grad_norm": 3.796875,
+      "learning_rate": 9.766659409232918e-06,
+      "loss": 2.1032,
+      "step": 108
+    },
+    {
+      "epoch": 0.14433262711864406,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.759999146340031e-06,
+      "loss": 2.0202,
+      "step": 109
+    },
+    {
+      "epoch": 0.14565677966101695,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.753247499983649e-06,
+      "loss": 2.0955,
+      "step": 110
+    },
+    {
+      "epoch": 0.14698093220338984,
+      "grad_norm": 4.03125,
+      "learning_rate": 9.7464045997833e-06,
+      "loss": 2.019,
+      "step": 111
+    },
+    {
+      "epoch": 0.1483050847457627,
+      "grad_norm": 4.625,
+      "learning_rate": 9.739470577110417e-06,
+      "loss": 2.1751,
+      "step": 112
+    },
+    {
+      "epoch": 0.1496292372881356,
+      "grad_norm": 4.125,
+      "learning_rate": 9.732445565085823e-06,
+      "loss": 2.0698,
+      "step": 113
+    },
+    {
+      "epoch": 0.15095338983050846,
+      "grad_norm": 4.03125,
+      "learning_rate": 9.725329698577177e-06,
+      "loss": 2.1954,
+      "step": 114
+    },
+    {
+      "epoch": 0.15227754237288135,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.718123114196381e-06,
+      "loss": 2.1319,
+      "step": 115
+    },
+    {
+      "epoch": 0.15360169491525424,
+      "grad_norm": 3.921875,
+      "learning_rate": 9.71082595029695e-06,
+      "loss": 2.0517,
+      "step": 116
+    },
+    {
+      "epoch": 0.1549258474576271,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.703438346971373e-06,
+      "loss": 2.243,
+      "step": 117
+    },
+    {
+      "epoch": 0.15625,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.69596044604841e-06,
+      "loss": 2.2454,
+      "step": 118
+    },
+    {
+      "epoch": 0.1575741525423729,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.688392391090374e-06,
+      "loss": 2.0891,
+      "step": 119
+    },
+    {
+      "epoch": 0.15889830508474576,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.680734327390374e-06,
+      "loss": 2.1075,
+      "step": 120
+    },
+    {
+      "epoch": 0.16022245762711865,
+      "grad_norm": 4.125,
+      "learning_rate": 9.672986401969523e-06,
+      "loss": 2.0766,
+      "step": 121
+    },
+    {
+      "epoch": 0.16154661016949154,
+      "grad_norm": 3.71875,
+      "learning_rate": 9.665148763574123e-06,
+      "loss": 1.9083,
+      "step": 122
+    },
+    {
+      "epoch": 0.1628707627118644,
+      "grad_norm": 3.9375,
+      "learning_rate": 9.657221562672803e-06,
+      "loss": 1.9674,
+      "step": 123
+    },
+    {
+      "epoch": 0.1641949152542373,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.64920495145363e-06,
+      "loss": 2.2366,
+      "step": 124
+    },
+    {
+      "epoch": 0.16551906779661016,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.64109908382119e-06,
+      "loss": 2.1767,
+      "step": 125
+    },
+    {
+      "epoch": 0.16684322033898305,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.632904115393633e-06,
+      "loss": 2.0395,
+      "step": 126
+    },
+    {
+      "epoch": 0.16816737288135594,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.624620203499683e-06,
+      "loss": 2.2272,
+      "step": 127
+    },
+    {
+      "epoch": 0.1694915254237288,
+      "grad_norm": 3.8125,
+      "learning_rate": 9.616247507175624e-06,
+      "loss": 2.094,
+      "step": 128
+    },
+    {
+      "epoch": 0.1708156779661017,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.607786187162234e-06,
+      "loss": 2.1253,
+      "step": 129
+    },
+    {
+      "epoch": 0.1721398305084746,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.599236405901715e-06,
+      "loss": 2.2173,
+      "step": 130
+    },
+    {
+      "epoch": 0.17346398305084745,
+      "grad_norm": 3.90625,
+      "learning_rate": 9.590598327534563e-06,
+      "loss": 2.0337,
+      "step": 131
+    },
+    {
+      "epoch": 0.17478813559322035,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.581872117896423e-06,
+      "loss": 2.0791,
+      "step": 132
+    },
+    {
+      "epoch": 0.1761122881355932,
+      "grad_norm": 3.859375,
+      "learning_rate": 9.573057944514897e-06,
+      "loss": 2.0358,
+      "step": 133
+    },
+    {
+      "epoch": 0.1774364406779661,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.56415597660634e-06,
+      "loss": 1.8235,
+      "step": 134
+    },
+    {
+      "epoch": 0.178760593220339,
+      "grad_norm": 4.46875,
+      "learning_rate": 9.555166385072599e-06,
+      "loss": 2.154,
+      "step": 135
+    },
+    {
+      "epoch": 0.18008474576271186,
+      "grad_norm": 3.953125,
+      "learning_rate": 9.546089342497743e-06,
+      "loss": 2.1079,
+      "step": 136
+    },
+    {
+      "epoch": 0.18140889830508475,
+      "grad_norm": 3.84375,
+      "learning_rate": 9.536925023144742e-06,
+      "loss": 2.1117,
+      "step": 137
+    },
+    {
+      "epoch": 0.18273305084745764,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.527673602952123e-06,
+      "loss": 2.0338,
+      "step": 138
+    },
+    {
+      "epoch": 0.1840572033898305,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.518335259530594e-06,
+      "loss": 2.2235,
+      "step": 139
+    },
+    {
+      "epoch": 0.1853813559322034,
+      "grad_norm": 4.84375,
+      "learning_rate": 9.508910172159635e-06,
+      "loss": 1.9275,
+      "step": 140
+    },
+    {
+      "epoch": 0.18670550847457626,
+      "grad_norm": 4.25,
+      "learning_rate": 9.499398521784051e-06,
+      "loss": 2.0549,
+      "step": 141
+    },
+    {
+      "epoch": 0.18802966101694915,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.489800491010507e-06,
+      "loss": 1.9433,
+      "step": 142
+    },
+    {
+      "epoch": 0.18935381355932204,
+      "grad_norm": 3.96875,
+      "learning_rate": 9.48011626410401e-06,
+      "loss": 2.2499,
+      "step": 143
+    },
+    {
+      "epoch": 0.1906779661016949,
+      "grad_norm": 3.8125,
+      "learning_rate": 9.470346026984387e-06,
+      "loss": 2.1125,
+      "step": 144
+    },
+    {
+      "epoch": 0.1920021186440678,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.4604899672227e-06,
+      "loss": 2.1017,
+      "step": 145
+    },
+    {
+      "epoch": 0.1933262711864407,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.450548274037652e-06,
+      "loss": 2.038,
+      "step": 146
+    },
+    {
+      "epoch": 0.19465042372881355,
+      "grad_norm": 4.375,
+      "learning_rate": 9.440521138291962e-06,
+      "loss": 2.1801,
+      "step": 147
+    },
+    {
+      "epoch": 0.19597457627118645,
+      "grad_norm": 3.734375,
+      "learning_rate": 9.430408752488687e-06,
+      "loss": 2.1859,
+      "step": 148
+    },
+    {
+      "epoch": 0.1972987288135593,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.420211310767534e-06,
+      "loss": 2.074,
+      "step": 149
+    },
+    {
+      "epoch": 0.1986228813559322,
+      "grad_norm": 4.8125,
+      "learning_rate": 9.409929008901126e-06,
+      "loss": 2.067,
+      "step": 150
+    },
+    {
+      "epoch": 0.1999470338983051,
+      "grad_norm": 3.671875,
+      "learning_rate": 9.399562044291261e-06,
+      "loss": 1.9712,
+      "step": 151
+    },
+    {
+      "epoch": 0.20127118644067796,
+      "grad_norm": 3.96875,
+      "learning_rate": 9.389110615965102e-06,
+      "loss": 2.0511,
+      "step": 152
+    },
+    {
+      "epoch": 0.20259533898305085,
+      "grad_norm": 4.0,
+      "learning_rate": 9.378574924571362e-06,
+      "loss": 2.0218,
+      "step": 153
+    },
+    {
+      "epoch": 0.20391949152542374,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.367955172376462e-06,
+      "loss": 1.9664,
+      "step": 154
+    },
+    {
+      "epoch": 0.2052436440677966,
+      "grad_norm": 3.859375,
+      "learning_rate": 9.35725156326063e-06,
+      "loss": 2.0343,
+      "step": 155
+    },
+    {
+      "epoch": 0.2065677966101695,
+      "grad_norm": 3.984375,
+      "learning_rate": 9.346464302714008e-06,
+      "loss": 2.2163,
+      "step": 156
+    },
+    {
+      "epoch": 0.20789194915254236,
+      "grad_norm": 3.609375,
+      "learning_rate": 9.335593597832686e-06,
+      "loss": 2.2098,
+      "step": 157
+    },
+    {
+      "epoch": 0.20921610169491525,
+      "grad_norm": 3.90625,
+      "learning_rate": 9.324639657314742e-06,
+      "loss": 2.0961,
+      "step": 158
+    },
+    {
+      "epoch": 0.21054025423728814,
+      "grad_norm": 3.984375,
+      "learning_rate": 9.313602691456224e-06,
+      "loss": 2.1141,
+      "step": 159
+    },
+    {
+      "epoch": 0.211864406779661,
+      "grad_norm": 3.96875,
+      "learning_rate": 9.302482912147126e-06,
+      "loss": 2.0155,
+      "step": 160
+    },
+    {
+      "epoch": 0.2131885593220339,
+      "grad_norm": 3.953125,
+      "learning_rate": 9.291280532867301e-06,
+      "loss": 1.9736,
+      "step": 161
+    },
+    {
+      "epoch": 0.2145127118644068,
+      "grad_norm": 3.734375,
+      "learning_rate": 9.279995768682383e-06,
+      "loss": 2.0853,
+      "step": 162
+    },
+    {
+      "epoch": 0.21583686440677965,
+      "grad_norm": 4.125,
+      "learning_rate": 9.268628836239646e-06,
+      "loss": 2.1189,
+      "step": 163
+    },
+    {
+      "epoch": 0.21716101694915255,
+      "grad_norm": 3.796875,
+      "learning_rate": 9.257179953763846e-06,
+      "loss": 2.0597,
+      "step": 164
+    },
+    {
+      "epoch": 0.2184851694915254,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.245649341053033e-06,
+      "loss": 2.0406,
+      "step": 165
+    },
+    {
+      "epoch": 0.2198093220338983,
+      "grad_norm": 3.96875,
+      "learning_rate": 9.234037219474332e-06,
+      "loss": 2.0314,
+      "step": 166
+    },
+    {
+      "epoch": 0.2211334745762712,
+      "grad_norm": 4.0,
+      "learning_rate": 9.222343811959694e-06,
+      "loss": 2.1392,
+      "step": 167
+    },
+    {
+      "epoch": 0.22245762711864406,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.21056934300161e-06,
+      "loss": 2.0291,
+      "step": 168
+    },
+    {
+      "epoch": 0.22378177966101695,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.198714038648811e-06,
+      "loss": 2.1251,
+      "step": 169
+    },
+    {
+      "epoch": 0.22510593220338984,
+      "grad_norm": 4.25,
+      "learning_rate": 9.186778126501916e-06,
+      "loss": 1.9912,
+      "step": 170
+    },
+    {
+      "epoch": 0.2264300847457627,
+      "grad_norm": 4.0,
+      "learning_rate": 9.17476183570908e-06,
+      "loss": 2.0754,
+      "step": 171
+    },
+    {
+      "epoch": 0.2277542372881356,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.162665396961573e-06,
+      "loss": 2.0376,
+      "step": 172
+    },
+    {
+      "epoch": 0.22907838983050846,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.150489042489368e-06,
+      "loss": 1.984,
+      "step": 173
+    },
+    {
+      "epoch": 0.23040254237288135,
+      "grad_norm": 3.984375,
+      "learning_rate": 9.138233006056679e-06,
+      "loss": 2.0155,
+      "step": 174
+    },
+    {
+      "epoch": 0.23172669491525424,
+      "grad_norm": 4.125,
+      "learning_rate": 9.125897522957461e-06,
+      "loss": 2.1034,
+      "step": 175
+    },
+    {
+      "epoch": 0.2330508474576271,
+      "grad_norm": 3.84375,
+      "learning_rate": 9.113482830010918e-06,
+      "loss": 2.1072,
+      "step": 176
+    },
+    {
+      "epoch": 0.234375,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.100989165556928e-06,
+      "loss": 2.1256,
+      "step": 177
+    },
+    {
+      "epoch": 0.2356991525423729,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.088416769451485e-06,
+      "loss": 2.0479,
+      "step": 178
+    },
+    {
+      "epoch": 0.23702330508474576,
+      "grad_norm": 4.0,
+      "learning_rate": 9.075765883062093e-06,
+      "loss": 2.075,
+      "step": 179
+    },
+    {
+      "epoch": 0.23834745762711865,
+      "grad_norm": 3.8125,
+      "learning_rate": 9.063036749263127e-06,
+      "loss": 1.9578,
+      "step": 180
+    },
+    {
+      "epoch": 0.23967161016949154,
+      "grad_norm": 3.796875,
+      "learning_rate": 9.050229612431168e-06,
+      "loss": 2.2268,
+      "step": 181
+    },
+    {
+      "epoch": 0.2409957627118644,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.037344718440321e-06,
+      "loss": 2.2065,
+      "step": 182
+    },
+    {
+      "epoch": 0.2423199152542373,
+      "grad_norm": 4.75,
+      "learning_rate": 9.02438231465749e-06,
+      "loss": 1.9338,
+      "step": 183
+    },
+    {
+      "epoch": 0.24364406779661016,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.011342649937623e-06,
+      "loss": 2.0992,
+      "step": 184
+    },
+    {
+      "epoch": 0.24496822033898305,
+      "grad_norm": 4.625,
+      "learning_rate": 8.99822597461894e-06,
+      "loss": 2.0273,
+      "step": 185
+    },
+    {
+      "epoch": 0.24629237288135594,
+      "grad_norm": 4.59375,
+      "learning_rate": 8.985032540518133e-06,
+      "loss": 2.0654,
+      "step": 186
+    },
+    {
+      "epoch": 0.2476165254237288,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.971762600925519e-06,
+      "loss": 2.1202,
+      "step": 187
+    },
+    {
+      "epoch": 0.2489406779661017,
+      "grad_norm": 3.75,
+      "learning_rate": 8.958416410600188e-06,
+      "loss": 2.0589,
+      "step": 188
+    },
+    {
+      "epoch": 0.2502648305084746,
+      "grad_norm": 4.0625,
+      "learning_rate": 8.944994225765104e-06,
+      "loss": 1.9044,
+      "step": 189
+    },
+    {
+      "epoch": 0.2515889830508475,
+      "grad_norm": 4.375,
+      "learning_rate": 8.931496304102192e-06,
+      "loss": 2.0083,
+      "step": 190
+    },
+    {
+      "epoch": 0.2529131355932203,
+      "grad_norm": 3.859375,
+      "learning_rate": 8.917922904747385e-06,
+      "loss": 1.9487,
+      "step": 191
+    },
+    {
+      "epoch": 0.2542372881355932,
+      "grad_norm": 3.734375,
+      "learning_rate": 8.904274288285657e-06,
+      "loss": 1.9682,
+      "step": 192
+    },
+    {
+      "epoch": 0.2555614406779661,
+      "grad_norm": 4.21875,
+      "learning_rate": 8.890550716746013e-06,
+      "loss": 1.8554,
+      "step": 193
+    },
+    {
+      "epoch": 0.256885593220339,
+      "grad_norm": 3.921875,
+      "learning_rate": 8.876752453596462e-06,
+      "loss": 2.0785,
+      "step": 194
+    },
+    {
+      "epoch": 0.2582097457627119,
+      "grad_norm": 3.796875,
+      "learning_rate": 8.862879763738962e-06,
+      "loss": 2.0384,
+      "step": 195
+    },
+    {
+      "epoch": 0.2595338983050847,
+      "grad_norm": 3.828125,
+      "learning_rate": 8.84893291350432e-06,
+      "loss": 2.163,
+      "step": 196
+    },
+    {
+      "epoch": 0.2608580508474576,
+      "grad_norm": 4.125,
+      "learning_rate": 8.834912170647102e-06,
+      "loss": 2.0237,
+      "step": 197
+    },
+    {
+      "epoch": 0.2621822033898305,
+      "grad_norm": 4.0625,
+      "learning_rate": 8.820817804340471e-06,
+      "loss": 2.0215,
+      "step": 198
+    },
+    {
+      "epoch": 0.2635063559322034,
+      "grad_norm": 3.671875,
+      "learning_rate": 8.806650085171036e-06,
+      "loss": 2.0454,
+      "step": 199
+    },
+    {
+      "epoch": 0.2648305084745763,
+      "grad_norm": 3.78125,
+      "learning_rate": 8.792409285133644e-06,
+      "loss": 2.0006,
+      "step": 200
+    },
+    {
+      "epoch": 0.2648305084745763,
+      "eval_AskDoctor-Chinese_loss": 2.021162986755371,
+      "eval_AskDoctor-Chinese_runtime": 16.0254,
+      "eval_AskDoctor-Chinese_samples_per_second": 1.872,
+      "eval_AskDoctor-Chinese_steps_per_second": 1.872,
+      "step": 200
+    },
+    {
+      "epoch": 0.2661546610169492,
+      "grad_norm": 3.921875,
+      "learning_rate": 8.778095677626164e-06,
+      "loss": 1.9454,
+      "step": 201
+    },
+    {
+      "epoch": 0.267478813559322,
+      "grad_norm": 3.765625,
+      "learning_rate": 8.763709537444241e-06,
+      "loss": 2.1055,
+      "step": 202
+    },
+    {
+      "epoch": 0.2688029661016949,
+      "grad_norm": 4.59375,
+      "learning_rate": 8.749251140776016e-06,
+      "loss": 1.9506,
+      "step": 203
+    },
+    {
+      "epoch": 0.2701271186440678,
+      "grad_norm": 3.9375,
+      "learning_rate": 8.734720765196826e-06,
+      "loss": 2.2382,
+      "step": 204
+    },
+    {
+      "epoch": 0.2714512711864407,
+      "grad_norm": 4.96875,
+      "learning_rate": 8.720118689663872e-06,
+      "loss": 1.9558,
+      "step": 205
+    },
+    {
+      "epoch": 0.2727754237288136,
+      "grad_norm": 3.84375,
+      "learning_rate": 8.705445194510868e-06,
+      "loss": 2.0589,
+      "step": 206
+    },
+    {
+      "epoch": 0.2740995762711864,
+      "grad_norm": 4.84375,
+      "learning_rate": 8.690700561442658e-06,
+      "loss": 1.8821,
+      "step": 207
+    },
+    {
+      "epoch": 0.2754237288135593,
+      "grad_norm": 3.84375,
+      "learning_rate": 8.675885073529802e-06,
+      "loss": 1.9863,
+      "step": 208
+    },
+    {
+      "epoch": 0.2767478813559322,
+      "grad_norm": 3.84375,
+      "learning_rate": 8.660999015203152e-06,
+      "loss": 2.0464,
+      "step": 209
+    },
+    {
+      "epoch": 0.2780720338983051,
+      "grad_norm": 3.9375,
+      "learning_rate": 8.64604267224838e-06,
+      "loss": 2.1284,
+      "step": 210
+    },
+    {
+      "epoch": 0.279396186440678,
+      "grad_norm": 4.21875,
+      "learning_rate": 8.631016331800501e-06,
+      "loss": 1.9782,
+      "step": 211
+    },
+    {
+      "epoch": 0.2807203389830508,
+      "grad_norm": 3.84375,
+      "learning_rate": 8.615920282338355e-06,
+      "loss": 2.0504,
+      "step": 212
+    },
+    {
+      "epoch": 0.2820444915254237,
+      "grad_norm": 3.859375,
+      "learning_rate": 8.600754813679072e-06,
+      "loss": 2.0384,
+      "step": 213
+    },
+    {
+      "epoch": 0.2833686440677966,
+      "grad_norm": 4.1875,
+      "learning_rate": 8.585520216972503e-06,
+      "loss": 2.0057,
+      "step": 214
+    },
+    {
+      "epoch": 0.2846927966101695,
+      "grad_norm": 3.765625,
+      "learning_rate": 8.570216784695637e-06,
+      "loss": 2.1141,
+      "step": 215
+    },
+    {
+      "epoch": 0.2860169491525424,
+      "grad_norm": 4.1875,
+      "learning_rate": 8.55484481064698e-06,
+      "loss": 1.9975,
+      "step": 216
+    },
+    {
+      "epoch": 0.2873411016949153,
+      "grad_norm": 4.03125,
+      "learning_rate": 8.539404589940924e-06,
+      "loss": 1.9758,
+      "step": 217
+    },
+    {
+      "epoch": 0.2886652542372881,
+      "grad_norm": 6.0625,
+      "learning_rate": 8.52389641900206e-06,
+      "loss": 1.8848,
+      "step": 218
+    },
+    {
+      "epoch": 0.289989406779661,
+      "grad_norm": 3.734375,
+      "learning_rate": 8.50832059555952e-06,
+      "loss": 2.09,
+      "step": 219
+    },
+    {
+      "epoch": 0.2913135593220339,
+      "grad_norm": 3.984375,
+      "learning_rate": 8.492677418641231e-06,
+      "loss": 1.824,
+      "step": 220
+    },
+    {
+      "epoch": 0.2926377118644068,
+      "grad_norm": 3.90625,
+      "learning_rate": 8.476967188568187e-06,
+      "loss": 2.1443,
+      "step": 221
+    },
+    {
+      "epoch": 0.2939618644067797,
+      "grad_norm": 3.796875,
+      "learning_rate": 8.461190206948691e-06,
+      "loss": 2.1238,
+      "step": 222
+    },
+    {
+      "epoch": 0.2952860169491525,
+      "grad_norm": 4.0625,
+      "learning_rate": 8.445346776672546e-06,
+      "loss": 2.1187,
+      "step": 223
+    },
+    {
+      "epoch": 0.2966101694915254,
+      "grad_norm": 3.65625,
+      "learning_rate": 8.429437201905254e-06,
+      "loss": 2.2054,
+      "step": 224
+    },
+    {
+      "epoch": 0.2979343220338983,
+      "grad_norm": 3.5,
+      "learning_rate": 8.413461788082175e-06,
+      "loss": 2.0788,
+      "step": 225
+    },
+    {
+      "epoch": 0.2992584745762712,
+      "grad_norm": 3.5625,
+      "learning_rate": 8.39742084190266e-06,
+      "loss": 2.1441,
+      "step": 226
+    },
+    {
+      "epoch": 0.3005826271186441,
+      "grad_norm": 7.3125,
+      "learning_rate": 8.38131467132416e-06,
+      "loss": 1.9429,
+      "step": 227
+    },
+    {
+      "epoch": 0.3019067796610169,
+      "grad_norm": 7.375,
+      "learning_rate": 8.365143585556326e-06,
+      "loss": 1.9466,
+      "step": 228
+    },
+    {
+      "epoch": 0.3032309322033898,
+      "grad_norm": 4.625,
+      "learning_rate": 8.348907895055055e-06,
+      "loss": 1.9793,
+      "step": 229
+    },
+    {
+      "epoch": 0.3045550847457627,
+      "grad_norm": 4.09375,
+      "learning_rate": 8.332607911516545e-06,
+      "loss": 2.0855,
+      "step": 230
+    },
+    {
+      "epoch": 0.3058792372881356,
+      "grad_norm": 4.125,
+      "learning_rate": 8.316243947871306e-06,
+      "loss": 2.1595,
+      "step": 231
+    },
+    {
+      "epoch": 0.3072033898305085,
+      "grad_norm": 3.953125,
+      "learning_rate": 8.299816318278146e-06,
+      "loss": 1.9467,
+      "step": 232
+    },
+    {
+      "epoch": 0.3085275423728814,
+      "grad_norm": 3.75,
+      "learning_rate": 8.283325338118154e-06,
+      "loss": 1.9687,
+      "step": 233
+    },
+    {
+      "epoch": 0.3098516949152542,
+      "grad_norm": 3.875,
+      "learning_rate": 8.266771323988624e-06,
+      "loss": 2.0397,
+      "step": 234
+    },
+    {
+      "epoch": 0.3111758474576271,
+      "grad_norm": 3.890625,
+      "learning_rate": 8.250154593697002e-06,
+      "loss": 2.1581,
+      "step": 235
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 3.640625,
+      "learning_rate": 8.233475466254766e-06,
+      "loss": 1.9687,
+      "step": 236
+    },
+    {
+      "epoch": 0.3138241525423729,
+      "grad_norm": 3.828125,
+      "learning_rate": 8.216734261871305e-06,
+      "loss": 2.081,
+      "step": 237
+    },
+    {
+      "epoch": 0.3151483050847458,
+      "grad_norm": 3.984375,
+      "learning_rate": 8.199931301947782e-06,
+      "loss": 1.9637,
+      "step": 238
+    },
+    {
+      "epoch": 0.3164724576271186,
+      "grad_norm": 6.96875,
+      "learning_rate": 8.183066909070946e-06,
+      "loss": 1.9271,
+      "step": 239
+    },
+    {
+      "epoch": 0.3177966101694915,
+      "grad_norm": 3.9375,
+      "learning_rate": 8.16614140700696e-06,
+      "loss": 1.9426,
+      "step": 240
+    },
+    {
+      "epoch": 0.3191207627118644,
+      "grad_norm": 4.03125,
+      "learning_rate": 8.149155120695163e-06,
+      "loss": 1.8951,
+      "step": 241
+    },
+    {
+      "epoch": 0.3204449152542373,
+      "grad_norm": 3.84375,
+      "learning_rate": 8.132108376241849e-06,
+      "loss": 2.1779,
+      "step": 242
+    },
+    {
+      "epoch": 0.3217690677966102,
+      "grad_norm": 3.921875,
+      "learning_rate": 8.115001500914e-06,
+      "loss": 2.1942,
+      "step": 243
+    },
+    {
+      "epoch": 0.3230932203389831,
+      "grad_norm": 3.90625,
+      "learning_rate": 8.097834823133002e-06,
+      "loss": 1.9011,
+      "step": 244
+    },
+    {
+      "epoch": 0.3244173728813559,
+      "grad_norm": 4.15625,
+      "learning_rate": 8.08060867246834e-06,
+      "loss": 1.9409,
+      "step": 245
+    },
+    {
+      "epoch": 0.3257415254237288,
+      "grad_norm": 4.96875,
+      "learning_rate": 8.063323379631274e-06,
+      "loss": 1.9138,
+      "step": 246
+    },
+    {
+      "epoch": 0.3270656779661017,
+      "grad_norm": 3.875,
+      "learning_rate": 8.045979276468486e-06,
+      "loss": 2.182,
+      "step": 247
+    },
+    {
+      "epoch": 0.3283898305084746,
+      "grad_norm": 3.90625,
+      "learning_rate": 8.028576695955711e-06,
+      "loss": 2.0034,
+      "step": 248
+    },
+    {
+      "epoch": 0.3297139830508475,
+      "grad_norm": 3.6875,
+      "learning_rate": 8.011115972191347e-06,
+      "loss": 2.1779,
+      "step": 249
+    },
+    {
+      "epoch": 0.3310381355932203,
+      "grad_norm": 5.25,
+      "learning_rate": 7.993597440390035e-06,
+      "loss": 1.8994,
+      "step": 250
+    },
+    {
+      "epoch": 0.3323622881355932,
+      "grad_norm": 3.765625,
+      "learning_rate": 7.976021436876232e-06,
+      "loss": 2.0865,
+      "step": 251
+    },
+    {
+      "epoch": 0.3336864406779661,
+      "grad_norm": 3.90625,
+      "learning_rate": 7.958388299077739e-06,
+      "loss": 2.1092,
+      "step": 252
+    },
+    {
+      "epoch": 0.335010593220339,
+      "grad_norm": 4.0,
+      "learning_rate": 7.940698365519246e-06,
+      "loss": 2.0966,
+      "step": 253
+    },
+    {
+      "epoch": 0.3363347457627119,
+      "grad_norm": 3.625,
+      "learning_rate": 7.92295197581581e-06,
+      "loss": 2.0216,
+      "step": 254
+    },
+    {
+      "epoch": 0.3376588983050847,
+      "grad_norm": 3.796875,
+      "learning_rate": 7.905149470666348e-06,
+      "loss": 2.036,
+      "step": 255
+    },
+    {
+      "epoch": 0.3389830508474576,
+      "grad_norm": 4.625,
+      "learning_rate": 7.887291191847098e-06,
+      "loss": 2.0037,
+      "step": 256
+    },
+    {
+      "epoch": 0.3403072033898305,
+      "grad_norm": 4.0625,
+      "learning_rate": 7.869377482205042e-06,
+      "loss": 2.0473,
+      "step": 257
+    },
+    {
+      "epoch": 0.3416313559322034,
+      "grad_norm": 4.125,
+      "learning_rate": 7.851408685651342e-06,
+      "loss": 2.0279,
+      "step": 258
+    },
+    {
+      "epoch": 0.3429555084745763,
+      "grad_norm": 3.9375,
+      "learning_rate": 7.833385147154733e-06,
+      "loss": 2.1209,
+      "step": 259
+    },
+    {
+      "epoch": 0.3442796610169492,
+      "grad_norm": 3.96875,
+      "learning_rate": 7.815307212734888e-06,
+      "loss": 2.0669,
+      "step": 260
+    },
+    {
+      "epoch": 0.345603813559322,
+      "grad_norm": 3.71875,
+      "learning_rate": 7.797175229455793e-06,
+      "loss": 2.033,
+      "step": 261
+    },
+    {
+      "epoch": 0.3469279661016949,
+      "grad_norm": 4.46875,
+      "learning_rate": 7.778989545419068e-06,
+      "loss": 2.0184,
+      "step": 262
+    },
+    {
+      "epoch": 0.3482521186440678,
+      "grad_norm": 3.765625,
+      "learning_rate": 7.7607505097573e-06,
+      "loss": 2.2567,
+      "step": 263
+    },
+    {
+      "epoch": 0.3495762711864407,
+      "grad_norm": 3.6875,
+      "learning_rate": 7.742458472627321e-06,
+      "loss": 2.0145,
+      "step": 264
+    },
+    {
+      "epoch": 0.3509004237288136,
+      "grad_norm": 3.953125,
+      "learning_rate": 7.724113785203506e-06,
+      "loss": 1.94,
+      "step": 265
+    },
+    {
+      "epoch": 0.3522245762711864,
+      "grad_norm": 3.78125,
+      "learning_rate": 7.705716799671019e-06,
+      "loss": 2.0266,
+      "step": 266
+    },
+    {
+      "epoch": 0.3535487288135593,
+      "grad_norm": 3.953125,
+      "learning_rate": 7.687267869219052e-06,
+      "loss": 1.9502,
+      "step": 267
+    },
+    {
+      "epoch": 0.3548728813559322,
+      "grad_norm": 3.953125,
+      "learning_rate": 7.668767348034044e-06,
+      "loss": 2.1102,
+      "step": 268
+    },
+    {
+      "epoch": 0.3561970338983051,
+      "grad_norm": 3.90625,
+      "learning_rate": 7.650215591292888e-06,
+      "loss": 1.8979,
+      "step": 269
+    },
+    {
+      "epoch": 0.357521186440678,
+      "grad_norm": 4.125,
+      "learning_rate": 7.631612955156111e-06,
+      "loss": 1.9134,
+      "step": 270
+    },
+    {
+      "epoch": 0.3588453389830508,
+      "grad_norm": 3.84375,
+      "learning_rate": 7.6129597967610244e-06,
+      "loss": 2.0089,
+      "step": 271
+    },
+    {
+      "epoch": 0.3601694915254237,
+      "grad_norm": 3.890625,
+      "learning_rate": 7.594256474214883e-06,
+      "loss": 1.9536,
+      "step": 272
+    },
+    {
+      "epoch": 0.3614936440677966,
+      "grad_norm": 3.984375,
+      "learning_rate": 7.5755033465880024e-06,
+      "loss": 2.0965,
+      "step": 273
+    },
+    {
+      "epoch": 0.3628177966101695,
+      "grad_norm": 4.28125,
+      "learning_rate": 7.556700773906866e-06,
+      "loss": 2.0323,
+      "step": 274
+    },
+    {
+      "epoch": 0.3641419491525424,
+      "grad_norm": 3.8125,
+      "learning_rate": 7.537849117147212e-06,
+      "loss": 2.0372,
+      "step": 275
+    },
+    {
+      "epoch": 0.3654661016949153,
+      "grad_norm": 3.640625,
+      "learning_rate": 7.5189487382271095e-06,
+      "loss": 2.0669,
+      "step": 276
+    },
+    {
+      "epoch": 0.3667902542372881,
+      "grad_norm": 3.703125,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 1.9949,
+      "step": 277
+    },
+    {
+      "epoch": 0.368114406779661,
+      "grad_norm": 3.703125,
+      "learning_rate": 7.481003266247745e-06,
+      "loss": 2.1448,
+      "step": 278
+    },
+    {
+      "epoch": 0.3694385593220339,
+      "grad_norm": 3.90625,
+      "learning_rate": 7.461958901673625e-06,
+      "loss": 2.0343,
+      "step": 279
+    },
+    {
+      "epoch": 0.3707627118644068,
+      "grad_norm": 4.25,
+      "learning_rate": 7.4428672718953535e-06,
+      "loss": 2.0603,
+      "step": 280
+    },
+    {
+      "epoch": 0.3720868644067797,
+      "grad_norm": 3.640625,
+      "learning_rate": 7.4237287434380485e-06,
+      "loss": 1.9779,
+      "step": 281
+    },
+    {
+      "epoch": 0.3734110169491525,
+      "grad_norm": 3.78125,
+      "learning_rate": 7.404543683727201e-06,
+      "loss": 2.1302,
+      "step": 282
+    },
+    {
+      "epoch": 0.3747351694915254,
+      "grad_norm": 6.15625,
+      "learning_rate": 7.385312461081616e-06,
+      "loss": 1.9677,
+      "step": 283
+    },
+    {
+      "epoch": 0.3760593220338983,
+      "grad_norm": 7.6875,
+      "learning_rate": 7.366035444706346e-06,
+      "loss": 1.969,
+      "step": 284
+    },
+    {
+      "epoch": 0.3773834745762712,
+      "grad_norm": 3.953125,
+      "learning_rate": 7.346713004685602e-06,
+      "loss": 2.0783,
+      "step": 285
+    },
+    {
+      "epoch": 0.3787076271186441,
+      "grad_norm": 3.875,
+      "learning_rate": 7.3273455119756445e-06,
+      "loss": 2.134,
+      "step": 286
+    },
+    {
+      "epoch": 0.3800317796610169,
+      "grad_norm": 4.125,
+      "learning_rate": 7.307933338397667e-06,
+      "loss": 2.0337,
+      "step": 287
+    },
+    {
+      "epoch": 0.3813559322033898,
+      "grad_norm": 3.59375,
+      "learning_rate": 7.288476856630656e-06,
+      "loss": 2.2021,
+      "step": 288
+    },
+    {
+      "epoch": 0.3826800847457627,
+      "grad_norm": 3.765625,
+      "learning_rate": 7.268976440204236e-06,
+      "loss": 2.0299,
+      "step": 289
+    },
+    {
+      "epoch": 0.3840042372881356,
+      "grad_norm": 4.15625,
+      "learning_rate": 7.249432463491498e-06,
+      "loss": 1.9574,
+      "step": 290
+    },
+    {
+      "epoch": 0.3853283898305085,
+      "grad_norm": 3.734375,
+      "learning_rate": 7.229845301701811e-06,
+      "loss": 2.2027,
+      "step": 291
+    },
+    {
+      "epoch": 0.3866525423728814,
+      "grad_norm": 3.578125,
+      "learning_rate": 7.2102153308736225e-06,
+      "loss": 2.1389,
+      "step": 292
+    },
+    {
+      "epoch": 0.3879766949152542,
+      "grad_norm": 3.796875,
+      "learning_rate": 7.190542927867234e-06,
+      "loss": 1.8233,
+      "step": 293
+    },
+    {
+      "epoch": 0.3893008474576271,
+      "grad_norm": 3.984375,
+      "learning_rate": 7.1708284703575734e-06,
+      "loss": 1.9075,
+      "step": 294
+    },
+    {
+      "epoch": 0.390625,
+      "grad_norm": 4.5625,
+      "learning_rate": 7.1510723368269376e-06,
+      "loss": 1.8944,
+      "step": 295
+    },
+    {
+      "epoch": 0.3919491525423729,
+      "grad_norm": 4.0,
+      "learning_rate": 7.131274906557725e-06,
+      "loss": 2.1448,
+      "step": 296
+    },
+    {
+      "epoch": 0.3932733050847458,
+      "grad_norm": 4.34375,
+      "learning_rate": 7.111436559625162e-06,
+      "loss": 2.0403,
+      "step": 297
+    },
+    {
+      "epoch": 0.3945974576271186,
+      "grad_norm": 3.90625,
+      "learning_rate": 7.091557676890001e-06,
+      "loss": 1.982,
+      "step": 298
+    },
+    {
+      "epoch": 0.3959216101694915,
+      "grad_norm": 4.5,
+      "learning_rate": 7.0716386399912075e-06,
+      "loss": 2.1055,
+      "step": 299
+    },
+    {
+      "epoch": 0.3972457627118644,
+      "grad_norm": 3.515625,
+      "learning_rate": 7.051679831338638e-06,
+      "loss": 1.8751,
+      "step": 300
+    },
+    {
+      "epoch": 0.3985699152542373,
+      "grad_norm": 3.765625,
+      "learning_rate": 7.03168163410569e-06,
+      "loss": 2.0518,
+      "step": 301
+    },
+    {
+      "epoch": 0.3998940677966102,
+      "grad_norm": 3.8125,
+      "learning_rate": 7.0116444322219575e-06,
+      "loss": 1.9739,
+      "step": 302
+    },
+    {
+      "epoch": 0.4012182203389831,
+      "grad_norm": 3.578125,
+      "learning_rate": 6.991568610365851e-06,
+      "loss": 2.0927,
+      "step": 303
+    },
+    {
+      "epoch": 0.4025423728813559,
+      "grad_norm": 4.4375,
+      "learning_rate": 6.971454553957216e-06,
+      "loss": 2.0478,
+      "step": 304
+    },
+    {
+      "epoch": 0.4038665254237288,
+      "grad_norm": 4.15625,
+      "learning_rate": 6.95130264914993e-06,
+      "loss": 2.0242,
+      "step": 305
+    },
+    {
+      "epoch": 0.4051906779661017,
+      "grad_norm": 3.65625,
+      "learning_rate": 6.931113282824498e-06,
+      "loss": 2.1597,
+      "step": 306
+    },
+    {
+      "epoch": 0.4065148305084746,
+      "grad_norm": 3.734375,
+      "learning_rate": 6.910886842580612e-06,
+      "loss": 1.9689,
+      "step": 307
+    },
+    {
+      "epoch": 0.4078389830508475,
+      "grad_norm": 3.640625,
+      "learning_rate": 6.890623716729724e-06,
+      "loss": 2.1267,
+      "step": 308
+    },
+    {
+      "epoch": 0.4091631355932203,
+      "grad_norm": 3.796875,
+      "learning_rate": 6.870324294287578e-06,
+      "loss": 2.0548,
+      "step": 309
+    },
+    {
+      "epoch": 0.4104872881355932,
+      "grad_norm": 3.703125,
+      "learning_rate": 6.8499889649667516e-06,
+      "loss": 2.2051,
+      "step": 310
+    },
+    {
+      "epoch": 0.4118114406779661,
+      "grad_norm": 3.765625,
+      "learning_rate": 6.829618119169169e-06,
+      "loss": 1.9724,
+      "step": 311
+    },
+    {
+      "epoch": 0.413135593220339,
+      "grad_norm": 4.09375,
+      "learning_rate": 6.809212147978605e-06,
+      "loss": 2.0979,
+      "step": 312
+    },
+    {
+      "epoch": 0.4144597457627119,
+      "grad_norm": 3.953125,
+      "learning_rate": 6.788771443153183e-06,
+      "loss": 2.1663,
+      "step": 313
+    },
+    {
+      "epoch": 0.4157838983050847,
+      "grad_norm": 4.03125,
+      "learning_rate": 6.768296397117848e-06,
+      "loss": 1.9478,
+      "step": 314
+    },
+    {
+      "epoch": 0.4171080508474576,
+      "grad_norm": 4.03125,
+      "learning_rate": 6.7477874029568345e-06,
+      "loss": 2.0767,
+      "step": 315
+    },
+    {
+      "epoch": 0.4184322033898305,
+      "grad_norm": 5.875,
+      "learning_rate": 6.7272448544061184e-06,
+      "loss": 2.0509,
+      "step": 316
+    },
+    {
+      "epoch": 0.4197563559322034,
+      "grad_norm": 3.921875,
+      "learning_rate": 6.706669145845863e-06,
+      "loss": 2.0612,
+      "step": 317
+    },
+    {
+      "epoch": 0.4210805084745763,
+      "grad_norm": 5.09375,
+      "learning_rate": 6.686060672292847e-06,
+      "loss": 1.9404,
+      "step": 318
+    },
+    {
+      "epoch": 0.4224046610169492,
+      "grad_norm": 4.0,
+      "learning_rate": 6.6654198293928695e-06,
+      "loss": 2.0939,
+      "step": 319
+    },
+    {
+      "epoch": 0.423728813559322,
+      "grad_norm": 4.84375,
+      "learning_rate": 6.6447470134131685e-06,
+      "loss": 1.9003,
+      "step": 320
+    },
+    {
+      "epoch": 0.4250529661016949,
+      "grad_norm": 4.0,
+      "learning_rate": 6.624042621234814e-06,
+      "loss": 2.1884,
+      "step": 321
+    },
+    {
+      "epoch": 0.4263771186440678,
+      "grad_norm": 3.734375,
+      "learning_rate": 6.603307050345069e-06,
+      "loss": 1.9816,
+      "step": 322
+    },
+    {
+      "epoch": 0.4277012711864407,
+      "grad_norm": 3.828125,
+      "learning_rate": 6.5825406988297815e-06,
+      "loss": 1.9909,
+      "step": 323
+    },
+    {
+      "epoch": 0.4290254237288136,
+      "grad_norm": 4.03125,
+      "learning_rate": 6.561743965365732e-06,
+      "loss": 2.1602,
+      "step": 324
+    },
+    {
+      "epoch": 0.4303495762711864,
+      "grad_norm": 3.625,
+      "learning_rate": 6.540917249212976e-06,
+      "loss": 2.2546,
+      "step": 325
+    },
+    {
+      "epoch": 0.4316737288135593,
+      "grad_norm": 3.734375,
+      "learning_rate": 6.520060950207186e-06,
+      "loss": 2.1109,
+      "step": 326
+    },
+    {
+      "epoch": 0.4329978813559322,
+      "grad_norm": 3.6875,
+      "learning_rate": 6.4991754687519695e-06,
+      "loss": 1.9424,
+      "step": 327
+    },
+    {
+      "epoch": 0.4343220338983051,
+      "grad_norm": 4.03125,
+      "learning_rate": 6.478261205811188e-06,
+      "loss": 1.9637,
+      "step": 328
+    },
+    {
+      "epoch": 0.435646186440678,
+      "grad_norm": 4.1875,
+      "learning_rate": 6.457318562901257e-06,
+      "loss": 1.9564,
+      "step": 329
+    },
+    {
+      "epoch": 0.4369703389830508,
+      "grad_norm": 3.734375,
+      "learning_rate": 6.43634794208343e-06,
+      "loss": 2.1901,
+      "step": 330
+    },
+    {
+      "epoch": 0.4382944915254237,
+      "grad_norm": 4.0625,
+      "learning_rate": 6.415349745956093e-06,
+      "loss": 1.8634,
+      "step": 331
+    },
+    {
+      "epoch": 0.4396186440677966,
+      "grad_norm": 4.1875,
+      "learning_rate": 6.394324377647028e-06,
+      "loss": 1.9514,
+      "step": 332
+    },
+    {
+      "epoch": 0.4409427966101695,
+      "grad_norm": 3.953125,
+      "learning_rate": 6.373272240805668e-06,
+      "loss": 1.8502,
+      "step": 333
+    },
+    {
+      "epoch": 0.4422669491525424,
+      "grad_norm": 3.78125,
+      "learning_rate": 6.35219373959536e-06,
+      "loss": 1.9593,
+      "step": 334
+    },
+    {
+      "epoch": 0.4435911016949153,
+      "grad_norm": 3.765625,
+      "learning_rate": 6.331089278685599e-06,
+      "loss": 2.057,
+      "step": 335
+    },
+    {
+      "epoch": 0.4449152542372881,
+      "grad_norm": 4.84375,
+      "learning_rate": 6.30995926324426e-06,
+      "loss": 1.9964,
+      "step": 336
+    },
+    {
+      "epoch": 0.446239406779661,
+      "grad_norm": 3.796875,
+      "learning_rate": 6.2888040989298136e-06,
+      "loss": 2.2322,
+      "step": 337
+    },
+    {
+      "epoch": 0.4475635593220339,
+      "grad_norm": 3.84375,
+      "learning_rate": 6.267624191883551e-06,
+      "loss": 1.9032,
+      "step": 338
+    },
+    {
+      "epoch": 0.4488877118644068,
+      "grad_norm": 3.90625,
+      "learning_rate": 6.246419948721777e-06,
+      "loss": 2.026,
+      "step": 339
+    },
+    {
+      "epoch": 0.4502118644067797,
+      "grad_norm": 3.625,
+      "learning_rate": 6.2251917765280056e-06,
+      "loss": 2.1106,
+      "step": 340
+    },
+    {
+      "epoch": 0.4515360169491525,
+      "grad_norm": 3.859375,
+      "learning_rate": 6.203940082845144e-06,
+      "loss": 1.8843,
+      "step": 341
+    },
+    {
+      "epoch": 0.4528601694915254,
+      "grad_norm": 4.0,
+      "learning_rate": 6.182665275667674e-06,
+      "loss": 1.8529,
+      "step": 342
+    },
+    {
+      "epoch": 0.4541843220338983,
+      "grad_norm": 3.859375,
+      "learning_rate": 6.161367763433812e-06,
+      "loss": 2.0598,
+      "step": 343
+    },
+    {
+      "epoch": 0.4555084745762712,
+      "grad_norm": 3.859375,
+      "learning_rate": 6.140047955017672e-06,
+      "loss": 2.0318,
+      "step": 344
+    },
+    {
+      "epoch": 0.4568326271186441,
+      "grad_norm": 4.09375,
+      "learning_rate": 6.118706259721414e-06,
+      "loss": 2.0361,
+      "step": 345
+    },
+    {
+      "epoch": 0.4581567796610169,
+      "grad_norm": 3.984375,
+      "learning_rate": 6.097343087267386e-06,
+      "loss": 2.0142,
+      "step": 346
+    },
+    {
+      "epoch": 0.4594809322033898,
+      "grad_norm": 4.09375,
+      "learning_rate": 6.075958847790262e-06,
+      "loss": 2.0354,
+      "step": 347
+    },
+    {
+      "epoch": 0.4608050847457627,
+      "grad_norm": 3.765625,
+      "learning_rate": 6.054553951829163e-06,
+      "loss": 1.9838,
+      "step": 348
+    },
+    {
+      "epoch": 0.4621292372881356,
+      "grad_norm": 4.03125,
+      "learning_rate": 6.033128810319779e-06,
+      "loss": 1.9332,
+      "step": 349
+    },
+    {
+      "epoch": 0.4634533898305085,
+      "grad_norm": 3.75,
+      "learning_rate": 6.011683834586474e-06,
+      "loss": 2.0284,
+      "step": 350
+    },
+    {
+      "epoch": 0.4647775423728814,
+      "grad_norm": 3.703125,
+      "learning_rate": 5.9902194363344014e-06,
+      "loss": 1.9933,
+      "step": 351
+    },
+    {
+      "epoch": 0.4661016949152542,
+      "grad_norm": 3.96875,
+      "learning_rate": 5.968736027641584e-06,
+      "loss": 1.9212,
+      "step": 352
+    },
+    {
+      "epoch": 0.4674258474576271,
+      "grad_norm": 4.3125,
+      "learning_rate": 5.947234020951015e-06,
+      "loss": 1.9715,
+      "step": 353
+    },
+    {
+      "epoch": 0.46875,
+      "grad_norm": 4.0625,
+      "learning_rate": 5.925713829062737e-06,
+      "loss": 2.116,
+      "step": 354
+    },
+    {
+      "epoch": 0.4700741525423729,
+      "grad_norm": 4.0,
+      "learning_rate": 5.904175865125915e-06,
+      "loss": 1.8785,
+      "step": 355
+    },
+    {
+      "epoch": 0.4713983050847458,
+      "grad_norm": 3.8125,
+      "learning_rate": 5.882620542630901e-06,
+      "loss": 2.1109,
+      "step": 356
+    },
+    {
+      "epoch": 0.4727224576271186,
+      "grad_norm": 4.0625,
+      "learning_rate": 5.86104827540131e-06,
+      "loss": 2.046,
+      "step": 357
+    },
+    {
+      "epoch": 0.4740466101694915,
+      "grad_norm": 3.890625,
+      "learning_rate": 5.839459477586056e-06,
+      "loss": 1.931,
+      "step": 358
+    },
+    {
+      "epoch": 0.4753707627118644,
+      "grad_norm": 4.0,
+      "learning_rate": 5.817854563651415e-06,
+      "loss": 1.9138,
+      "step": 359
+    },
+    {
+      "epoch": 0.4766949152542373,
+      "grad_norm": 3.90625,
+      "learning_rate": 5.796233948373061e-06,
+      "loss": 2.0626,
+      "step": 360
+    },
+    {
+      "epoch": 0.4780190677966102,
+      "grad_norm": 3.6875,
+      "learning_rate": 5.7745980468281116e-06,
+      "loss": 1.9901,
+      "step": 361
+    },
+    {
+      "epoch": 0.4793432203389831,
+      "grad_norm": 3.515625,
+      "learning_rate": 5.752947274387147e-06,
+      "loss": 2.1562,
+      "step": 362
+    },
+    {
+      "epoch": 0.4806673728813559,
+      "grad_norm": 4.03125,
+      "learning_rate": 5.731282046706247e-06,
+      "loss": 1.9464,
+      "step": 363
+    },
+    {
+      "epoch": 0.4819915254237288,
+      "grad_norm": 3.875,
+      "learning_rate": 5.709602779718999e-06,
+      "loss": 1.9471,
+      "step": 364
+    },
+    {
+      "epoch": 0.4833156779661017,
+      "grad_norm": 3.640625,
+      "learning_rate": 5.687909889628529e-06,
+      "loss": 1.9786,
+      "step": 365
+    },
+    {
+      "epoch": 0.4846398305084746,
+      "grad_norm": 3.796875,
+      "learning_rate": 5.666203792899496e-06,
+      "loss": 2.0612,
+      "step": 366
+    },
+    {
+      "epoch": 0.4859639830508475,
+      "grad_norm": 3.84375,
+      "learning_rate": 5.644484906250104e-06,
+      "loss": 1.9507,
+      "step": 367
+    },
+    {
+      "epoch": 0.4872881355932203,
+      "grad_norm": 3.984375,
+      "learning_rate": 5.622753646644102e-06,
+      "loss": 1.8828,
+      "step": 368
+    },
+    {
+      "epoch": 0.4886122881355932,
+      "grad_norm": 3.671875,
+      "learning_rate": 5.601010431282777e-06,
+      "loss": 1.9527,
+      "step": 369
+    },
+    {
+      "epoch": 0.4899364406779661,
+      "grad_norm": 3.90625,
+      "learning_rate": 5.579255677596944e-06,
+      "loss": 2.0724,
+      "step": 370
+    },
+    {
+      "epoch": 0.491260593220339,
+      "grad_norm": 3.890625,
+      "learning_rate": 5.557489803238934e-06,
+      "loss": 1.9829,
+      "step": 371
+    },
+    {
+      "epoch": 0.4925847457627119,
+      "grad_norm": 3.75,
+      "learning_rate": 5.535713226074576e-06,
+      "loss": 1.9163,
+      "step": 372
+    },
+    {
+      "epoch": 0.4939088983050847,
+      "grad_norm": 3.90625,
+      "learning_rate": 5.513926364175172e-06,
+      "loss": 2.0671,
+      "step": 373
+    },
+    {
+      "epoch": 0.4952330508474576,
+      "grad_norm": 3.75,
+      "learning_rate": 5.492129635809473e-06,
+      "loss": 2.0903,
+      "step": 374
+    },
+    {
+      "epoch": 0.4965572033898305,
+      "grad_norm": 4.125,
+      "learning_rate": 5.47032345943565e-06,
+      "loss": 2.0251,
+      "step": 375
+    },
+    {
+      "epoch": 0.4978813559322034,
+      "grad_norm": 3.65625,
+      "learning_rate": 5.4485082536932564e-06,
+      "loss": 2.0114,
+      "step": 376
+    },
+    {
+      "epoch": 0.4992055084745763,
+      "grad_norm": 3.734375,
+      "learning_rate": 5.426684437395196e-06,
+      "loss": 1.9574,
+      "step": 377
+    },
+    {
+      "epoch": 0.5005296610169492,
+      "grad_norm": 4.25,
+      "learning_rate": 5.404852429519678e-06,
+      "loss": 1.9259,
+      "step": 378
+    },
+    {
+      "epoch": 0.501853813559322,
+      "grad_norm": 4.03125,
+      "learning_rate": 5.383012649202173e-06,
+      "loss": 1.9558,
+      "step": 379
+    },
+    {
+      "epoch": 0.503177966101695,
+      "grad_norm": 3.96875,
+      "learning_rate": 5.361165515727374e-06,
+      "loss": 2.0474,
+      "step": 380
+    },
+    {
+      "epoch": 0.5045021186440678,
+      "grad_norm": 3.890625,
+      "learning_rate": 5.3393114485211394e-06,
+      "loss": 2.0343,
+      "step": 381
+    },
+    {
+      "epoch": 0.5058262711864406,
+      "grad_norm": 3.84375,
+      "learning_rate": 5.31745086714244e-06,
+      "loss": 1.9331,
+      "step": 382
+    },
+    {
+      "epoch": 0.5071504237288136,
+      "grad_norm": 4.0,
+      "learning_rate": 5.295584191275308e-06,
+      "loss": 1.8745,
+      "step": 383
+    },
+    {
+      "epoch": 0.5084745762711864,
+      "grad_norm": 3.8125,
+      "learning_rate": 5.273711840720783e-06,
+      "loss": 1.9222,
+      "step": 384
+    },
+    {
+      "epoch": 0.5097987288135594,
+      "grad_norm": 4.0,
+      "learning_rate": 5.251834235388845e-06,
+      "loss": 2.1023,
+      "step": 385
+    },
+    {
+      "epoch": 0.5111228813559322,
+      "grad_norm": 3.96875,
+      "learning_rate": 5.229951795290353e-06,
+      "loss": 1.9507,
+      "step": 386
+    },
+    {
+      "epoch": 0.512447033898305,
+      "grad_norm": 3.75,
+      "learning_rate": 5.208064940528994e-06,
+      "loss": 2.0536,
+      "step": 387
+    },
+    {
+      "epoch": 0.513771186440678,
+      "grad_norm": 3.6875,
+      "learning_rate": 5.1861740912932e-06,
+      "loss": 2.0979,
+      "step": 388
+    },
+    {
+      "epoch": 0.5150953389830508,
+      "grad_norm": 3.75,
+      "learning_rate": 5.164279667848094e-06,
+      "loss": 2.0845,
+      "step": 389
+    },
+    {
+      "epoch": 0.5164194915254238,
+      "grad_norm": 4.09375,
+      "learning_rate": 5.142382090527415e-06,
+      "loss": 2.0602,
+      "step": 390
+    },
+    {
+      "epoch": 0.5177436440677966,
+      "grad_norm": 3.953125,
+      "learning_rate": 5.1204817797254545e-06,
+      "loss": 1.883,
+      "step": 391
+    },
+    {
+      "epoch": 0.5190677966101694,
+      "grad_norm": 3.75,
+      "learning_rate": 5.0985791558889785e-06,
+      "loss": 2.131,
+      "step": 392
+    },
+    {
+      "epoch": 0.5203919491525424,
+      "grad_norm": 3.90625,
+      "learning_rate": 5.07667463950916e-06,
+      "loss": 2.0166,
+      "step": 393
+    },
+    {
+      "epoch": 0.5217161016949152,
+      "grad_norm": 3.703125,
+      "learning_rate": 5.054768651113506e-06,
+      "loss": 1.9136,
+      "step": 394
+    },
+    {
+      "epoch": 0.5230402542372882,
+      "grad_norm": 4.15625,
+      "learning_rate": 5.032861611257783e-06,
+      "loss": 2.0923,
+      "step": 395
+    },
+    {
+      "epoch": 0.524364406779661,
+      "grad_norm": 3.8125,
+      "learning_rate": 5.0109539405179456e-06,
+      "loss": 2.1243,
+      "step": 396
+    },
+    {
+      "epoch": 0.5256885593220338,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.989046059482055e-06,
+      "loss": 2.0784,
+      "step": 397
+    },
+    {
+      "epoch": 0.5270127118644068,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.967138388742218e-06,
+      "loss": 1.8999,
+      "step": 398
+    },
+    {
+      "epoch": 0.5283368644067796,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.945231348886495e-06,
+      "loss": 2.0906,
+      "step": 399
+    },
+    {
+      "epoch": 0.5296610169491526,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.923325360490841e-06,
+      "loss": 1.9295,
+      "step": 400
+    },
+    {
+      "epoch": 0.5296610169491526,
+      "eval_AskDoctor-Chinese_loss": 1.9834351539611816,
+      "eval_AskDoctor-Chinese_runtime": 15.9925,
+      "eval_AskDoctor-Chinese_samples_per_second": 1.876,
+      "eval_AskDoctor-Chinese_steps_per_second": 1.876,
+      "step": 400
+    },
+    {
+      "epoch": 0.5309851694915254,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.9014208441110215e-06,
+      "loss": 2.0197,
+      "step": 401
+    },
+    {
+      "epoch": 0.5323093220338984,
+      "grad_norm": 3.75,
+      "learning_rate": 4.879518220274546e-06,
+      "loss": 2.0886,
+      "step": 402
+    },
+    {
+      "epoch": 0.5336334745762712,
+      "grad_norm": 4.5,
+      "learning_rate": 4.8576179094725855e-06,
+      "loss": 1.9252,
+      "step": 403
+    },
+    {
+      "epoch": 0.534957627118644,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.835720332151907e-06,
+      "loss": 2.0754,
+      "step": 404
+    },
+    {
+      "epoch": 0.536281779661017,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.813825908706802e-06,
+      "loss": 2.1356,
+      "step": 405
+    },
+    {
+      "epoch": 0.5376059322033898,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.791935059471007e-06,
+      "loss": 2.0131,
+      "step": 406
+    },
+    {
+      "epoch": 0.5389300847457628,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.770048204709648e-06,
+      "loss": 1.8858,
+      "step": 407
+    },
+    {
+      "epoch": 0.5402542372881356,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.748165764611157e-06,
+      "loss": 1.9881,
+      "step": 408
+    },
+    {
+      "epoch": 0.5415783898305084,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.726288159279218e-06,
+      "loss": 2.0027,
+      "step": 409
+    },
+    {
+      "epoch": 0.5429025423728814,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.7044158087246926e-06,
+      "loss": 2.1449,
+      "step": 410
+    },
+    {
+      "epoch": 0.5442266949152542,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.682549132857562e-06,
+      "loss": 2.0456,
+      "step": 411
+    },
+    {
+      "epoch": 0.5455508474576272,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.660688551478861e-06,
+      "loss": 2.2308,
+      "step": 412
+    },
+    {
+      "epoch": 0.546875,
+      "grad_norm": 3.875,
+      "learning_rate": 4.6388344842726266e-06,
+      "loss": 2.0734,
+      "step": 413
+    },
+    {
+      "epoch": 0.5481991525423728,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.616987350797827e-06,
+      "loss": 2.0188,
+      "step": 414
+    },
+    {
+      "epoch": 0.5495233050847458,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.595147570480324e-06,
+      "loss": 2.1403,
+      "step": 415
+    },
+    {
+      "epoch": 0.5508474576271186,
+      "grad_norm": 5.75,
+      "learning_rate": 4.573315562604804e-06,
+      "loss": 1.8386,
+      "step": 416
+    },
+    {
+      "epoch": 0.5521716101694916,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.551491746306744e-06,
+      "loss": 2.0066,
+      "step": 417
+    },
+    {
+      "epoch": 0.5534957627118644,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.529676540564351e-06,
+      "loss": 1.9468,
+      "step": 418
+    },
+    {
+      "epoch": 0.5548199152542372,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.5078703641905275e-06,
+      "loss": 2.0946,
+      "step": 419
+    },
+    {
+      "epoch": 0.5561440677966102,
+      "grad_norm": 3.625,
+      "learning_rate": 4.486073635824831e-06,
+      "loss": 2.206,
+      "step": 420
+    },
+    {
+      "epoch": 0.557468220338983,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.464286773925426e-06,
+      "loss": 1.977,
+      "step": 421
+    },
+    {
+      "epoch": 0.558792372881356,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.442510196761068e-06,
+      "loss": 1.8473,
+      "step": 422
+    },
+    {
+      "epoch": 0.5601165254237288,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.420744322403058e-06,
+      "loss": 2.1905,
+      "step": 423
+    },
+    {
+      "epoch": 0.5614406779661016,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.398989568717226e-06,
+      "loss": 2.0168,
+      "step": 424
+    },
+    {
+      "epoch": 0.5627648305084746,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.377246353355899e-06,
+      "loss": 1.9603,
+      "step": 425
+    },
+    {
+      "epoch": 0.5640889830508474,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.355515093749897e-06,
+      "loss": 1.9786,
+      "step": 426
+    },
+    {
+      "epoch": 0.5654131355932204,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.333796207100505e-06,
+      "loss": 1.9911,
+      "step": 427
+    },
+    {
+      "epoch": 0.5667372881355932,
+      "grad_norm": 5.25,
+      "learning_rate": 4.312090110371473e-06,
+      "loss": 1.9617,
+      "step": 428
+    },
+    {
+      "epoch": 0.5680614406779662,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.290397220281002e-06,
+      "loss": 1.9384,
+      "step": 429
+    },
+    {
+      "epoch": 0.569385593220339,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.268717953293755e-06,
+      "loss": 2.1731,
+      "step": 430
+    },
+    {
+      "epoch": 0.5707097457627118,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.247052725612853e-06,
+      "loss": 1.8694,
+      "step": 431
+    },
+    {
+      "epoch": 0.5720338983050848,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.22540195317189e-06,
+      "loss": 1.8741,
+      "step": 432
+    },
+    {
+      "epoch": 0.5733580508474576,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.203766051626939e-06,
+      "loss": 1.9796,
+      "step": 433
+    },
+    {
+      "epoch": 0.5746822033898306,
+      "grad_norm": 4.0,
+      "learning_rate": 4.182145436348587e-06,
+      "loss": 1.9154,
+      "step": 434
+    },
+    {
+      "epoch": 0.5760063559322034,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.160540522413947e-06,
+      "loss": 1.9181,
+      "step": 435
+    },
+    {
+      "epoch": 0.5773305084745762,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.138951724598692e-06,
+      "loss": 2.0523,
+      "step": 436
+    },
+    {
+      "epoch": 0.5786546610169492,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.1173794573691e-06,
+      "loss": 1.8367,
+      "step": 437
+    },
+    {
+      "epoch": 0.579978813559322,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.095824134874087e-06,
+      "loss": 2.0884,
+      "step": 438
+    },
+    {
+      "epoch": 0.581302966101695,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.074286170937265e-06,
+      "loss": 2.0347,
+      "step": 439
+    },
+    {
+      "epoch": 0.5826271186440678,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.052765979048986e-06,
+      "loss": 2.0196,
+      "step": 440
+    },
+    {
+      "epoch": 0.5839512711864406,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.031263972358419e-06,
+      "loss": 1.9662,
+      "step": 441
+    },
+    {
+      "epoch": 0.5852754237288136,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.009780563665601e-06,
+      "loss": 1.9913,
+      "step": 442
+    },
+    {
+      "epoch": 0.5865995762711864,
+      "grad_norm": 3.90625,
+      "learning_rate": 3.988316165413528e-06,
+      "loss": 2.0734,
+      "step": 443
+    },
+    {
+      "epoch": 0.5879237288135594,
+      "grad_norm": 3.71875,
+      "learning_rate": 3.966871189680223e-06,
+      "loss": 1.9153,
+      "step": 444
+    },
+    {
+      "epoch": 0.5892478813559322,
+      "grad_norm": 3.96875,
+      "learning_rate": 3.945446048170839e-06,
+      "loss": 1.9798,
+      "step": 445
+    },
+    {
+      "epoch": 0.590572033898305,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.924041152209739e-06,
+      "loss": 1.9039,
+      "step": 446
+    },
+    {
+      "epoch": 0.591896186440678,
+      "grad_norm": 4.4375,
+      "learning_rate": 3.902656912732616e-06,
+      "loss": 1.9863,
+      "step": 447
+    },
+    {
+      "epoch": 0.5932203389830508,
+      "grad_norm": 3.984375,
+      "learning_rate": 3.881293740278588e-06,
+      "loss": 1.9465,
+      "step": 448
+    },
+    {
+      "epoch": 0.5945444915254238,
+      "grad_norm": 3.71875,
+      "learning_rate": 3.859952044982329e-06,
+      "loss": 2.0645,
+      "step": 449
+    },
+    {
+      "epoch": 0.5958686440677966,
+      "grad_norm": 3.828125,
+      "learning_rate": 3.83863223656619e-06,
+      "loss": 2.0472,
+      "step": 450
+    },
+    {
+      "epoch": 0.5971927966101694,
+      "grad_norm": 3.671875,
+      "learning_rate": 3.8173347243323275e-06,
+      "loss": 1.8184,
+      "step": 451
+    },
+    {
+      "epoch": 0.5985169491525424,
+      "grad_norm": 4.0625,
+      "learning_rate": 3.7960599171548572e-06,
+      "loss": 1.9835,
+      "step": 452
+    },
+    {
+      "epoch": 0.5998411016949152,
+      "grad_norm": 3.671875,
+      "learning_rate": 3.774808223471996e-06,
+      "loss": 1.9712,
+      "step": 453
+    },
+    {
+      "epoch": 0.6011652542372882,
+      "grad_norm": 3.703125,
+      "learning_rate": 3.7535800512782254e-06,
+      "loss": 2.0324,
+      "step": 454
+    },
+    {
+      "epoch": 0.602489406779661,
+      "grad_norm": 3.59375,
+      "learning_rate": 3.732375808116451e-06,
+      "loss": 2.1901,
+      "step": 455
+    },
+    {
+      "epoch": 0.6038135593220338,
+      "grad_norm": 3.625,
+      "learning_rate": 3.711195901070188e-06,
+      "loss": 2.031,
+      "step": 456
+    },
+    {
+      "epoch": 0.6051377118644068,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.690040736755742e-06,
+      "loss": 2.0682,
+      "step": 457
+    },
+    {
+      "epoch": 0.6064618644067796,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.6689107213144025e-06,
+      "loss": 2.1191,
+      "step": 458
+    },
+    {
+      "epoch": 0.6077860169491526,
+      "grad_norm": 3.65625,
+      "learning_rate": 3.6478062604046406e-06,
+      "loss": 1.921,
+      "step": 459
+    },
+    {
+      "epoch": 0.6091101694915254,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.626727759194334e-06,
+      "loss": 2.0207,
+      "step": 460
+    },
+    {
+      "epoch": 0.6104343220338984,
+      "grad_norm": 3.59375,
+      "learning_rate": 3.6056756223529734e-06,
+      "loss": 2.1463,
+      "step": 461
+    },
+    {
+      "epoch": 0.6117584745762712,
+      "grad_norm": 4.1875,
+      "learning_rate": 3.5846502540439076e-06,
+      "loss": 1.9282,
+      "step": 462
+    },
+    {
+      "epoch": 0.613082627118644,
+      "grad_norm": 3.53125,
+      "learning_rate": 3.5636520579165704e-06,
+      "loss": 2.1905,
+      "step": 463
+    },
+    {
+      "epoch": 0.614406779661017,
+      "grad_norm": 3.515625,
+      "learning_rate": 3.542681437098745e-06,
+      "loss": 2.1518,
+      "step": 464
+    },
+    {
+      "epoch": 0.6157309322033898,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.5217387941888117e-06,
+      "loss": 2.0076,
+      "step": 465
+    },
+    {
+      "epoch": 0.6170550847457628,
+      "grad_norm": 3.765625,
+      "learning_rate": 3.5008245312480326e-06,
+      "loss": 2.1018,
+      "step": 466
+    },
+    {
+      "epoch": 0.6183792372881356,
+      "grad_norm": 4.09375,
+      "learning_rate": 3.479939049792817e-06,
+      "loss": 1.8781,
+      "step": 467
+    },
+    {
+      "epoch": 0.6197033898305084,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.4590827507870257e-06,
+      "loss": 2.0801,
+      "step": 468
+    },
+    {
+      "epoch": 0.6210275423728814,
+      "grad_norm": 4.375,
+      "learning_rate": 3.4382560346342707e-06,
+      "loss": 2.211,
+      "step": 469
+    },
+    {
+      "epoch": 0.6223516949152542,
+      "grad_norm": 3.828125,
+      "learning_rate": 3.4174593011702197e-06,
+      "loss": 1.8481,
+      "step": 470
+    },
+    {
+      "epoch": 0.6236758474576272,
+      "grad_norm": 3.671875,
+      "learning_rate": 3.396692949654933e-06,
+      "loss": 2.0414,
+      "step": 471
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 3.984375,
+      "learning_rate": 3.3759573787651877e-06,
+      "loss": 2.094,
+      "step": 472
+    },
+    {
+      "epoch": 0.6263241525423728,
+      "grad_norm": 3.734375,
+      "learning_rate": 3.3552529865868323e-06,
+      "loss": 2.0496,
+      "step": 473
+    },
+    {
+      "epoch": 0.6276483050847458,
+      "grad_norm": 3.84375,
+      "learning_rate": 3.3345801706071325e-06,
+      "loss": 2.113,
+      "step": 474
+    },
+    {
+      "epoch": 0.6289724576271186,
+      "grad_norm": 4.03125,
+      "learning_rate": 3.3139393277071554e-06,
+      "loss": 1.8577,
+      "step": 475
+    },
+    {
+      "epoch": 0.6302966101694916,
+      "grad_norm": 5.21875,
+      "learning_rate": 3.2933308541541365e-06,
+      "loss": 2.059,
+      "step": 476
+    },
+    {
+      "epoch": 0.6316207627118644,
+      "grad_norm": 3.5625,
+      "learning_rate": 3.2727551455938832e-06,
+      "loss": 2.1471,
+      "step": 477
+    },
+    {
+      "epoch": 0.6329449152542372,
+      "grad_norm": 3.953125,
+      "learning_rate": 3.252212597043167e-06,
+      "loss": 2.0538,
+      "step": 478
+    },
+    {
+      "epoch": 0.6342690677966102,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.2317036028821523e-06,
+      "loss": 1.9275,
+      "step": 479
+    },
+    {
+      "epoch": 0.635593220338983,
+      "grad_norm": 4.125,
+      "learning_rate": 3.2112285568468166e-06,
+      "loss": 1.9979,
+      "step": 480
+    },
+    {
+      "epoch": 0.636917372881356,
+      "grad_norm": 3.796875,
+      "learning_rate": 3.190787852021396e-06,
+      "loss": 2.1286,
+      "step": 481
+    },
+    {
+      "epoch": 0.6382415254237288,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.1703818808308327e-06,
+      "loss": 2.007,
+      "step": 482
+    },
+    {
+      "epoch": 0.6395656779661016,
+      "grad_norm": 3.859375,
+      "learning_rate": 3.1500110350332492e-06,
+      "loss": 1.9457,
+      "step": 483
+    },
+    {
+      "epoch": 0.6408898305084746,
+      "grad_norm": 4.09375,
+      "learning_rate": 3.1296757057124243e-06,
+      "loss": 1.9245,
+      "step": 484
+    },
+    {
+      "epoch": 0.6422139830508474,
+      "grad_norm": 4.125,
+      "learning_rate": 3.1093762832702775e-06,
+      "loss": 2.0545,
+      "step": 485
+    },
+    {
+      "epoch": 0.6435381355932204,
+      "grad_norm": 3.859375,
+      "learning_rate": 3.08911315741939e-06,
+      "loss": 2.0104,
+      "step": 486
+    },
+    {
+      "epoch": 0.6448622881355932,
+      "grad_norm": 3.515625,
+      "learning_rate": 3.0688867171755043e-06,
+      "loss": 2.0609,
+      "step": 487
+    },
+    {
+      "epoch": 0.6461864406779662,
+      "grad_norm": 3.609375,
+      "learning_rate": 3.048697350850073e-06,
+      "loss": 2.0671,
+      "step": 488
+    },
+    {
+      "epoch": 0.647510593220339,
+      "grad_norm": 3.96875,
+      "learning_rate": 3.0285454460427855e-06,
+      "loss": 1.9852,
+      "step": 489
+    },
+    {
+      "epoch": 0.6488347457627118,
+      "grad_norm": 3.859375,
+      "learning_rate": 3.0084313896341504e-06,
+      "loss": 1.9553,
+      "step": 490
+    },
+    {
+      "epoch": 0.6501588983050848,
+      "grad_norm": 3.625,
+      "learning_rate": 2.988355567778043e-06,
+      "loss": 2.1243,
+      "step": 491
+    },
+    {
+      "epoch": 0.6514830508474576,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.9683183658943106e-06,
+      "loss": 2.137,
+      "step": 492
+    },
+    {
+      "epoch": 0.6528072033898306,
+      "grad_norm": 3.9375,
+      "learning_rate": 2.9483201686613626e-06,
+      "loss": 1.9655,
+      "step": 493
+    },
+    {
+      "epoch": 0.6541313559322034,
+      "grad_norm": 4.03125,
+      "learning_rate": 2.9283613600087933e-06,
+      "loss": 2.0074,
+      "step": 494
+    },
+    {
+      "epoch": 0.6554555084745762,
+      "grad_norm": 3.84375,
+      "learning_rate": 2.9084423231099985e-06,
+      "loss": 1.8278,
+      "step": 495
+    },
+    {
+      "epoch": 0.6567796610169492,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.8885634403748397e-06,
+      "loss": 2.0682,
+      "step": 496
+    },
+    {
+      "epoch": 0.658103813559322,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.8687250934422774e-06,
+      "loss": 2.0537,
+      "step": 497
+    },
+    {
+      "epoch": 0.659427966101695,
+      "grad_norm": 3.828125,
+      "learning_rate": 2.8489276631730633e-06,
+      "loss": 1.848,
+      "step": 498
+    },
+    {
+      "epoch": 0.6607521186440678,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.8291715296424282e-06,
+      "loss": 1.9693,
+      "step": 499
+    },
+    {
+      "epoch": 0.6620762711864406,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.809457072132766e-06,
+      "loss": 1.8889,
+      "step": 500
+    },
+    {
+      "epoch": 0.6634004237288136,
+      "grad_norm": 4.0,
+      "learning_rate": 2.789784669126379e-06,
+      "loss": 1.9459,
+      "step": 501
+    },
+    {
+      "epoch": 0.6647245762711864,
+      "grad_norm": 4.15625,
+      "learning_rate": 2.7701546982981903e-06,
+      "loss": 1.8192,
+      "step": 502
+    },
+    {
+      "epoch": 0.6660487288135594,
+      "grad_norm": 4.28125,
+      "learning_rate": 2.750567536508504e-06,
+      "loss": 1.9726,
+      "step": 503
+    },
+    {
+      "epoch": 0.6673728813559322,
+      "grad_norm": 4.0,
+      "learning_rate": 2.7310235597957652e-06,
+      "loss": 1.9488,
+      "step": 504
+    },
+    {
+      "epoch": 0.668697033898305,
+      "grad_norm": 4.59375,
+      "learning_rate": 2.711523143369345e-06,
+      "loss": 1.9806,
+      "step": 505
+    },
+    {
+      "epoch": 0.670021186440678,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.692066661602333e-06,
+      "loss": 2.1243,
+      "step": 506
+    },
+    {
+      "epoch": 0.6713453389830508,
+      "grad_norm": 3.9375,
+      "learning_rate": 2.6726544880243576e-06,
+      "loss": 1.9533,
+      "step": 507
+    },
+    {
+      "epoch": 0.6726694915254238,
+      "grad_norm": 3.75,
+      "learning_rate": 2.653286995314398e-06,
+      "loss": 2.1302,
+      "step": 508
+    },
+    {
+      "epoch": 0.6739936440677966,
+      "grad_norm": 3.84375,
+      "learning_rate": 2.633964555293654e-06,
+      "loss": 1.8948,
+      "step": 509
+    },
+    {
+      "epoch": 0.6753177966101694,
+      "grad_norm": 4.21875,
+      "learning_rate": 2.6146875389183845e-06,
+      "loss": 1.9675,
+      "step": 510
+    },
+    {
+      "epoch": 0.6766419491525424,
+      "grad_norm": 3.8125,
+      "learning_rate": 2.5954563162727996e-06,
+      "loss": 1.9404,
+      "step": 511
+    },
+    {
+      "epoch": 0.6779661016949152,
+      "grad_norm": 3.859375,
+      "learning_rate": 2.576271256561953e-06,
+      "loss": 1.9558,
+      "step": 512
+    },
+    {
+      "epoch": 0.6792902542372882,
+      "grad_norm": 4.3125,
+      "learning_rate": 2.5571327281046486e-06,
+      "loss": 1.9642,
+      "step": 513
+    },
+    {
+      "epoch": 0.680614406779661,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.538041098326377e-06,
+      "loss": 2.0344,
+      "step": 514
+    },
+    {
+      "epoch": 0.6819385593220338,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.5189967337522574e-06,
+      "loss": 1.8918,
+      "step": 515
+    },
+    {
+      "epoch": 0.6832627118644068,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 2.0214,
+      "step": 516
+    },
+    {
+      "epoch": 0.6845868644067796,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.481051261772891e-06,
+      "loss": 2.1078,
+      "step": 517
+    },
+    {
+      "epoch": 0.6859110169491526,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.46215088285279e-06,
+      "loss": 2.0772,
+      "step": 518
+    },
+    {
+      "epoch": 0.6872351694915254,
+      "grad_norm": 3.828125,
+      "learning_rate": 2.443299226093135e-06,
+      "loss": 1.9595,
+      "step": 519
+    },
+    {
+      "epoch": 0.6885593220338984,
+      "grad_norm": 3.859375,
+      "learning_rate": 2.4244966534119984e-06,
+      "loss": 2.0241,
+      "step": 520
+    },
+    {
+      "epoch": 0.6898834745762712,
+      "grad_norm": 4.1875,
+      "learning_rate": 2.4057435257851173e-06,
+      "loss": 1.9407,
+      "step": 521
+    },
+    {
+      "epoch": 0.691207627118644,
+      "grad_norm": 4.0,
+      "learning_rate": 2.387040203238977e-06,
+      "loss": 2.0936,
+      "step": 522
+    },
+    {
+      "epoch": 0.692531779661017,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.3683870448438905e-06,
+      "loss": 2.0738,
+      "step": 523
+    },
+    {
+      "epoch": 0.6938559322033898,
+      "grad_norm": 3.84375,
+      "learning_rate": 2.349784408707112e-06,
+      "loss": 1.9615,
+      "step": 524
+    },
+    {
+      "epoch": 0.6951800847457628,
+      "grad_norm": 3.625,
+      "learning_rate": 2.331232651965956e-06,
+      "loss": 1.9273,
+      "step": 525
+    },
+    {
+      "epoch": 0.6965042372881356,
+      "grad_norm": 3.90625,
+      "learning_rate": 2.312732130780951e-06,
+      "loss": 1.9053,
+      "step": 526
+    },
+    {
+      "epoch": 0.6978283898305084,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.2942832003289823e-06,
+      "loss": 2.0261,
+      "step": 527
+    },
+    {
+      "epoch": 0.6991525423728814,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.2758862147964933e-06,
+      "loss": 2.1048,
+      "step": 528
+    },
+    {
+      "epoch": 0.7004766949152542,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.2575415273726813e-06,
+      "loss": 2.1657,
+      "step": 529
+    },
+    {
+      "epoch": 0.7018008474576272,
+      "grad_norm": 4.59375,
+      "learning_rate": 2.2392494902427027e-06,
+      "loss": 1.8836,
+      "step": 530
+    },
+    {
+      "epoch": 0.703125,
+      "grad_norm": 3.875,
+      "learning_rate": 2.2210104545809337e-06,
+      "loss": 2.122,
+      "step": 531
+    },
+    {
+      "epoch": 0.7044491525423728,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.2028247705442095e-06,
+      "loss": 1.9893,
+      "step": 532
+    },
+    {
+      "epoch": 0.7057733050847458,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.1846927872651135e-06,
+      "loss": 2.1539,
+      "step": 533
+    },
+    {
+      "epoch": 0.7070974576271186,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.1666148528452692e-06,
+      "loss": 2.0765,
+      "step": 534
+    },
+    {
+      "epoch": 0.7084216101694916,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.1485913143486588e-06,
+      "loss": 2.0188,
+      "step": 535
+    },
+    {
+      "epoch": 0.7097457627118644,
+      "grad_norm": 3.90625,
+      "learning_rate": 2.1306225177949584e-06,
+      "loss": 1.9499,
+      "step": 536
+    },
+    {
+      "epoch": 0.7110699152542372,
+      "grad_norm": 3.9375,
+      "learning_rate": 2.1127088081529046e-06,
+      "loss": 1.9053,
+      "step": 537
+    },
+    {
+      "epoch": 0.7123940677966102,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.0948505293336506e-06,
+      "loss": 1.9854,
+      "step": 538
+    },
+    {
+      "epoch": 0.713718220338983,
+      "grad_norm": 3.921875,
+      "learning_rate": 2.07704802418419e-06,
+      "loss": 1.9253,
+      "step": 539
+    },
+    {
+      "epoch": 0.715042372881356,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.0593016344807544e-06,
+      "loss": 2.1557,
+      "step": 540
+    },
+    {
+      "epoch": 0.7163665254237288,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.0416117009222615e-06,
+      "loss": 2.0635,
+      "step": 541
+    },
+    {
+      "epoch": 0.7176906779661016,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.023978563123771e-06,
+      "loss": 2.0287,
+      "step": 542
+    },
+    {
+      "epoch": 0.7190148305084746,
+      "grad_norm": 3.9375,
+      "learning_rate": 2.0064025596099663e-06,
+      "loss": 1.8408,
+      "step": 543
+    },
+    {
+      "epoch": 0.7203389830508474,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.988884027808655e-06,
+      "loss": 1.9911,
+      "step": 544
+    },
+    {
+      "epoch": 0.7216631355932204,
+      "grad_norm": 3.875,
+      "learning_rate": 1.9714233040442915e-06,
+      "loss": 2.1129,
+      "step": 545
+    },
+    {
+      "epoch": 0.7229872881355932,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.954020723531517e-06,
+      "loss": 2.0941,
+      "step": 546
+    },
+    {
+      "epoch": 0.7243114406779662,
+      "grad_norm": 3.875,
+      "learning_rate": 1.9366766203687278e-06,
+      "loss": 2.0688,
+      "step": 547
+    },
+    {
+      "epoch": 0.725635593220339,
+      "grad_norm": 3.875,
+      "learning_rate": 1.919391327531663e-06,
+      "loss": 1.9566,
+      "step": 548
+    },
+    {
+      "epoch": 0.7269597457627118,
+      "grad_norm": 3.96875,
+      "learning_rate": 1.902165176866999e-06,
+      "loss": 2.2026,
+      "step": 549
+    },
+    {
+      "epoch": 0.7282838983050848,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.884998499086001e-06,
+      "loss": 2.047,
+      "step": 550
+    },
+    {
+      "epoch": 0.7296080508474576,
+      "grad_norm": 3.953125,
+      "learning_rate": 1.8678916237581524e-06,
+      "loss": 1.9537,
+      "step": 551
+    },
+    {
+      "epoch": 0.7309322033898306,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.8508448793048389e-06,
+      "loss": 1.9387,
+      "step": 552
+    },
+    {
+      "epoch": 0.7322563559322034,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.8338585929930424e-06,
+      "loss": 2.1303,
+      "step": 553
+    },
+    {
+      "epoch": 0.7335805084745762,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.816933090929055e-06,
+      "loss": 2.037,
+      "step": 554
+    },
+    {
+      "epoch": 0.7349046610169492,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.8000686980522193e-06,
+      "loss": 2.0009,
+      "step": 555
+    },
+    {
+      "epoch": 0.736228813559322,
+      "grad_norm": 3.984375,
+      "learning_rate": 1.7832657381286967e-06,
+      "loss": 2.0694,
+      "step": 556
+    },
+    {
+      "epoch": 0.737552966101695,
+      "grad_norm": 3.625,
+      "learning_rate": 1.7665245337452368e-06,
+      "loss": 2.1752,
+      "step": 557
+    },
+    {
+      "epoch": 0.7388771186440678,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.7498454063029984e-06,
+      "loss": 1.904,
+      "step": 558
+    },
+    {
+      "epoch": 0.7402012711864406,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.7332286760113782e-06,
+      "loss": 2.0804,
+      "step": 559
+    },
+    {
+      "epoch": 0.7415254237288136,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.716674661881848e-06,
+      "loss": 2.062,
+      "step": 560
+    },
+    {
+      "epoch": 0.7428495762711864,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.7001836817218542e-06,
+      "loss": 2.0387,
+      "step": 561
+    },
+    {
+      "epoch": 0.7441737288135594,
+      "grad_norm": 3.75,
+      "learning_rate": 1.6837560521286955e-06,
+      "loss": 1.9536,
+      "step": 562
+    },
+    {
+      "epoch": 0.7454978813559322,
+      "grad_norm": 3.84375,
+      "learning_rate": 1.667392088483456e-06,
+      "loss": 1.922,
+      "step": 563
+    },
+    {
+      "epoch": 0.746822033898305,
+      "grad_norm": 6.34375,
+      "learning_rate": 1.6510921049449469e-06,
+      "loss": 1.9446,
+      "step": 564
+    },
+    {
+      "epoch": 0.748146186440678,
+      "grad_norm": 3.96875,
+      "learning_rate": 1.6348564144436768e-06,
+      "loss": 1.9487,
+      "step": 565
+    },
+    {
+      "epoch": 0.7494703389830508,
+      "grad_norm": 3.75,
+      "learning_rate": 1.6186853286758397e-06,
+      "loss": 1.9243,
+      "step": 566
+    },
+    {
+      "epoch": 0.7507944915254238,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.6025791580973433e-06,
+      "loss": 2.0987,
+      "step": 567
+    },
+    {
+      "epoch": 0.7521186440677966,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.5865382119178258e-06,
+      "loss": 1.8775,
+      "step": 568
+    },
+    {
+      "epoch": 0.7534427966101694,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.570562798094747e-06,
+      "loss": 1.9614,
+      "step": 569
+    },
+    {
+      "epoch": 0.7547669491525424,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.554653223327456e-06,
+      "loss": 2.0446,
+      "step": 570
+    },
+    {
+      "epoch": 0.7560911016949152,
+      "grad_norm": 4.0,
+      "learning_rate": 1.5388097930513101e-06,
+      "loss": 2.0517,
+      "step": 571
+    },
+    {
+      "epoch": 0.7574152542372882,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.5230328114318127e-06,
+      "loss": 2.0772,
+      "step": 572
+    },
+    {
+      "epoch": 0.758739406779661,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.507322581358771e-06,
+      "loss": 1.9274,
+      "step": 573
+    },
+    {
+      "epoch": 0.7600635593220338,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.4916794044404814e-06,
+      "loss": 2.1604,
+      "step": 574
+    },
+    {
+      "epoch": 0.7613877118644068,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.4761035809979395e-06,
+      "loss": 1.9925,
+      "step": 575
+    },
+    {
+      "epoch": 0.7627118644067796,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.4605954100590796e-06,
+      "loss": 1.9394,
+      "step": 576
+    },
+    {
+      "epoch": 0.7640360169491526,
+      "grad_norm": 4.125,
+      "learning_rate": 1.4451551893530197e-06,
+      "loss": 1.9246,
+      "step": 577
+    },
+    {
+      "epoch": 0.7653601694915254,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.4297832153043657e-06,
+      "loss": 1.9387,
+      "step": 578
+    },
+    {
+      "epoch": 0.7666843220338984,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.414479783027498e-06,
+      "loss": 2.053,
+      "step": 579
+    },
+    {
+      "epoch": 0.7680084745762712,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.39924518632093e-06,
+      "loss": 1.9092,
+      "step": 580
+    },
+    {
+      "epoch": 0.769332627118644,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.3840797176616467e-06,
+      "loss": 2.048,
+      "step": 581
+    },
+    {
+      "epoch": 0.770656779661017,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.368983668199501e-06,
+      "loss": 1.897,
+      "step": 582
+    },
+    {
+      "epoch": 0.7719809322033898,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.353957327751621e-06,
+      "loss": 2.1015,
+      "step": 583
+    },
+    {
+      "epoch": 0.7733050847457628,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.3390009847968505e-06,
+      "loss": 1.9635,
+      "step": 584
+    },
+    {
+      "epoch": 0.7746292372881356,
+      "grad_norm": 4.28125,
+      "learning_rate": 1.3241149264701987e-06,
+      "loss": 1.9074,
+      "step": 585
+    },
+    {
+      "epoch": 0.7759533898305084,
+      "grad_norm": 4.75,
+      "learning_rate": 1.3092994385573432e-06,
+      "loss": 2.139,
+      "step": 586
+    },
+    {
+      "epoch": 0.7772775423728814,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.2945548054891322e-06,
+      "loss": 1.973,
+      "step": 587
+    },
+    {
+      "epoch": 0.7786016949152542,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.2798813103361291e-06,
+      "loss": 2.1142,
+      "step": 588
+    },
+    {
+      "epoch": 0.7799258474576272,
+      "grad_norm": 3.875,
+      "learning_rate": 1.2652792348031755e-06,
+      "loss": 2.2209,
+      "step": 589
+    },
+    {
+      "epoch": 0.78125,
+      "grad_norm": 4.21875,
+      "learning_rate": 1.2507488592239848e-06,
+      "loss": 1.9521,
+      "step": 590
+    },
+    {
+      "epoch": 0.7825741525423728,
+      "grad_norm": 4.96875,
+      "learning_rate": 1.2362904625557604e-06,
+      "loss": 1.8734,
+      "step": 591
+    },
+    {
+      "epoch": 0.7838983050847458,
+      "grad_norm": 3.625,
+      "learning_rate": 1.2219043223738375e-06,
+      "loss": 2.0146,
+      "step": 592
+    },
+    {
+      "epoch": 0.7852224576271186,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.2075907148663579e-06,
+      "loss": 1.9868,
+      "step": 593
+    },
+    {
+      "epoch": 0.7865466101694916,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.193349914828964e-06,
+      "loss": 2.1206,
+      "step": 594
+    },
+    {
+      "epoch": 0.7878707627118644,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.17918219565953e-06,
+      "loss": 2.0613,
+      "step": 595
+    },
+    {
+      "epoch": 0.7891949152542372,
+      "grad_norm": 4.125,
+      "learning_rate": 1.1650878293528994e-06,
+      "loss": 1.9358,
+      "step": 596
+    },
+    {
+      "epoch": 0.7905190677966102,
+      "grad_norm": 4.0,
+      "learning_rate": 1.1510670864956813e-06,
+      "loss": 1.8479,
+      "step": 597
+    },
+    {
+      "epoch": 0.791843220338983,
+      "grad_norm": 3.921875,
+      "learning_rate": 1.1371202362610412e-06,
+      "loss": 2.0899,
+      "step": 598
+    },
+    {
+      "epoch": 0.793167372881356,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.1232475464035386e-06,
+      "loss": 2.0043,
+      "step": 599
+    },
+    {
+      "epoch": 0.7944915254237288,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.1094492832539888e-06,
+      "loss": 2.0016,
+      "step": 600
+    },
+    {
+      "epoch": 0.7944915254237288,
+      "eval_AskDoctor-Chinese_loss": 1.9739471673965454,
+      "eval_AskDoctor-Chinese_runtime": 15.864,
+      "eval_AskDoctor-Chinese_samples_per_second": 1.891,
+      "eval_AskDoctor-Chinese_steps_per_second": 1.891,
+      "step": 600
+    },
+    {
+      "epoch": 0.7958156779661016,
+      "grad_norm": 3.5,
+      "learning_rate": 1.095725711714345e-06,
+      "loss": 2.1568,
+      "step": 601
+    },
+    {
+      "epoch": 0.7971398305084746,
+      "grad_norm": 3.75,
+      "learning_rate": 1.0820770952526155e-06,
+      "loss": 1.8976,
+      "step": 602
+    },
+    {
+      "epoch": 0.7984639830508474,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.06850369589781e-06,
+      "loss": 1.9435,
+      "step": 603
+    },
+    {
+      "epoch": 0.7997881355932204,
+      "grad_norm": 4.09375,
+      "learning_rate": 1.0550057742348973e-06,
+      "loss": 2.0421,
+      "step": 604
+    },
+    {
+      "epoch": 0.8011122881355932,
+      "grad_norm": 3.875,
+      "learning_rate": 1.0415835893998116e-06,
+      "loss": 1.9927,
+      "step": 605
+    },
+    {
+      "epoch": 0.8024364406779662,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.0282373990744821e-06,
+      "loss": 1.9352,
+      "step": 606
+    },
+    {
+      "epoch": 0.803760593220339,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.0149674594818676e-06,
+      "loss": 2.0455,
+      "step": 607
+    },
+    {
+      "epoch": 0.8050847457627118,
+      "grad_norm": 5.125,
+      "learning_rate": 1.0017740253810608e-06,
+      "loss": 1.9051,
+      "step": 608
+    },
+    {
+      "epoch": 0.8064088983050848,
+      "grad_norm": 3.53125,
+      "learning_rate": 9.886573500623791e-07,
+      "loss": 2.0362,
+      "step": 609
+    },
+    {
+      "epoch": 0.8077330508474576,
+      "grad_norm": 3.90625,
+      "learning_rate": 9.756176853425114e-07,
+      "loss": 2.0404,
+      "step": 610
+    },
+    {
+      "epoch": 0.8090572033898306,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.62655281559679e-07,
+      "loss": 2.0095,
+      "step": 611
+    },
+    {
+      "epoch": 0.8103813559322034,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.49770387568833e-07,
+      "loss": 1.896,
+      "step": 612
+    },
+    {
+      "epoch": 0.8117055084745762,
+      "grad_norm": 3.625,
+      "learning_rate": 9.369632507368736e-07,
+      "loss": 2.0449,
+      "step": 613
+    },
+    {
+      "epoch": 0.8130296610169492,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.242341169379077e-07,
+      "loss": 2.0941,
+      "step": 614
+    },
+    {
+      "epoch": 0.814353813559322,
+      "grad_norm": 3.75,
+      "learning_rate": 9.115832305485151e-07,
+      "loss": 2.0305,
+      "step": 615
+    },
+    {
+      "epoch": 0.815677966101695,
+      "grad_norm": 3.84375,
+      "learning_rate": 8.990108344430737e-07,
+      "loss": 1.8647,
+      "step": 616
+    },
+    {
+      "epoch": 0.8170021186440678,
+      "grad_norm": 3.8125,
+      "learning_rate": 8.865171699890835e-07,
+      "loss": 2.0226,
+      "step": 617
+    },
+    {
+      "epoch": 0.8183262711864406,
+      "grad_norm": 3.625,
+      "learning_rate": 8.741024770425394e-07,
+      "loss": 2.128,
+      "step": 618
+    },
+    {
+      "epoch": 0.8196504237288136,
+      "grad_norm": 3.53125,
+      "learning_rate": 8.617669939433237e-07,
+      "loss": 2.1359,
+      "step": 619
+    },
+    {
+      "epoch": 0.8209745762711864,
+      "grad_norm": 3.796875,
+      "learning_rate": 8.495109575106331e-07,
+      "loss": 2.0587,
+      "step": 620
+    },
+    {
+      "epoch": 0.8222987288135594,
+      "grad_norm": 3.828125,
+      "learning_rate": 8.373346030384288e-07,
+      "loss": 1.8764,
+      "step": 621
+    },
+    {
+      "epoch": 0.8236228813559322,
+      "grad_norm": 3.703125,
+      "learning_rate": 8.252381642909218e-07,
+      "loss": 2.0382,
+      "step": 622
+    },
+    {
+      "epoch": 0.824947033898305,
+      "grad_norm": 3.875,
+      "learning_rate": 8.132218734980852e-07,
+      "loss": 2.0033,
+      "step": 623
+    },
+    {
+      "epoch": 0.826271186440678,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.012859613511908e-07,
+      "loss": 1.9694,
+      "step": 624
+    },
+    {
+      "epoch": 0.8275953389830508,
+      "grad_norm": 3.734375,
+      "learning_rate": 7.894306569983923e-07,
+      "loss": 1.9737,
+      "step": 625
+    },
+    {
+      "epoch": 0.8289194915254238,
+      "grad_norm": 4.0,
+      "learning_rate": 7.776561880403072e-07,
+      "loss": 1.9066,
+      "step": 626
+    },
+    {
+      "epoch": 0.8302436440677966,
+      "grad_norm": 4.09375,
+      "learning_rate": 7.65962780525668e-07,
+      "loss": 1.913,
+      "step": 627
+    },
+    {
+      "epoch": 0.8315677966101694,
+      "grad_norm": 3.984375,
+      "learning_rate": 7.543506589469674e-07,
+      "loss": 1.8134,
+      "step": 628
+    },
+    {
+      "epoch": 0.8328919491525424,
+      "grad_norm": 3.65625,
+      "learning_rate": 7.42820046236154e-07,
+      "loss": 2.0282,
+      "step": 629
+    },
+    {
+      "epoch": 0.8342161016949152,
+      "grad_norm": 3.78125,
+      "learning_rate": 7.313711637603537e-07,
+      "loss": 1.985,
+      "step": 630
+    },
+    {
+      "epoch": 0.8355402542372882,
+      "grad_norm": 3.796875,
+      "learning_rate": 7.20004231317617e-07,
+      "loss": 2.0049,
+      "step": 631
+    },
+    {
+      "epoch": 0.836864406779661,
+      "grad_norm": 3.515625,
+      "learning_rate": 7.087194671326986e-07,
+      "loss": 2.0064,
+      "step": 632
+    },
+    {
+      "epoch": 0.8381885593220338,
+      "grad_norm": 3.828125,
+      "learning_rate": 6.975170878528765e-07,
+      "loss": 2.2509,
+      "step": 633
+    },
+    {
+      "epoch": 0.8395127118644068,
+      "grad_norm": 3.75,
+      "learning_rate": 6.863973085437769e-07,
+      "loss": 1.8559,
+      "step": 634
+    },
+    {
+      "epoch": 0.8408368644067796,
+      "grad_norm": 3.765625,
+      "learning_rate": 6.753603426852589e-07,
+      "loss": 1.9875,
+      "step": 635
+    },
+    {
+      "epoch": 0.8421610169491526,
+      "grad_norm": 3.78125,
+      "learning_rate": 6.644064021673158e-07,
+      "loss": 2.083,
+      "step": 636
+    },
+    {
+      "epoch": 0.8434851694915254,
+      "grad_norm": 3.90625,
+      "learning_rate": 6.53535697285993e-07,
+      "loss": 1.9355,
+      "step": 637
+    },
+    {
+      "epoch": 0.8448093220338984,
+      "grad_norm": 3.625,
+      "learning_rate": 6.427484367393699e-07,
+      "loss": 2.0545,
+      "step": 638
+    },
+    {
+      "epoch": 0.8461334745762712,
+      "grad_norm": 3.671875,
+      "learning_rate": 6.320448276235397e-07,
+      "loss": 2.0548,
+      "step": 639
+    },
+    {
+      "epoch": 0.847457627118644,
+      "grad_norm": 3.8125,
+      "learning_rate": 6.214250754286383e-07,
+      "loss": 2.1399,
+      "step": 640
+    },
+    {
+      "epoch": 0.848781779661017,
+      "grad_norm": 3.703125,
+      "learning_rate": 6.108893840348995e-07,
+      "loss": 2.1328,
+      "step": 641
+    },
+    {
+      "epoch": 0.8501059322033898,
+      "grad_norm": 3.9375,
+      "learning_rate": 6.004379557087397e-07,
+      "loss": 1.9439,
+      "step": 642
+    },
+    {
+      "epoch": 0.8514300847457628,
+      "grad_norm": 4.0,
+      "learning_rate": 5.900709910988739e-07,
+      "loss": 2.0008,
+      "step": 643
+    },
+    {
+      "epoch": 0.8527542372881356,
+      "grad_norm": 3.734375,
+      "learning_rate": 5.797886892324695e-07,
+      "loss": 1.9305,
+      "step": 644
+    },
+    {
+      "epoch": 0.8540783898305084,
+      "grad_norm": 3.765625,
+      "learning_rate": 5.695912475113136e-07,
+      "loss": 1.9696,
+      "step": 645
+    },
+    {
+      "epoch": 0.8554025423728814,
+      "grad_norm": 3.984375,
+      "learning_rate": 5.594788617080382e-07,
+      "loss": 2.0039,
+      "step": 646
+    },
+    {
+      "epoch": 0.8567266949152542,
+      "grad_norm": 4.15625,
+      "learning_rate": 5.494517259623478e-07,
+      "loss": 2.0625,
+      "step": 647
+    },
+    {
+      "epoch": 0.8580508474576272,
+      "grad_norm": 3.734375,
+      "learning_rate": 5.395100327773018e-07,
+      "loss": 2.1412,
+      "step": 648
+    },
+    {
+      "epoch": 0.859375,
+      "grad_norm": 8.6875,
+      "learning_rate": 5.296539730156142e-07,
+      "loss": 2.0818,
+      "step": 649
+    },
+    {
+      "epoch": 0.8606991525423728,
+      "grad_norm": 3.78125,
+      "learning_rate": 5.198837358959901e-07,
+      "loss": 2.0733,
+      "step": 650
+    },
+    {
+      "epoch": 0.8620233050847458,
+      "grad_norm": 3.9375,
+      "learning_rate": 5.101995089894946e-07,
+      "loss": 2.0582,
+      "step": 651
+    },
+    {
+      "epoch": 0.8633474576271186,
+      "grad_norm": 3.8125,
+      "learning_rate": 5.006014782159497e-07,
+      "loss": 2.0996,
+      "step": 652
+    },
+    {
+      "epoch": 0.8646716101694916,
+      "grad_norm": 3.5,
+      "learning_rate": 4.91089827840367e-07,
+      "loss": 2.1452,
+      "step": 653
+    },
+    {
+      "epoch": 0.8659957627118644,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.816647404694064e-07,
+      "loss": 1.9781,
+      "step": 654
+    },
+    {
+      "epoch": 0.8673199152542372,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.7232639704787864e-07,
+      "loss": 2.1196,
+      "step": 655
+    },
+    {
+      "epoch": 0.8686440677966102,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.6307497685525894e-07,
+      "loss": 2.1108,
+      "step": 656
+    },
+    {
+      "epoch": 0.869968220338983,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.5391065750225726e-07,
+      "loss": 1.966,
+      "step": 657
+    },
+    {
+      "epoch": 0.871292372881356,
+      "grad_norm": 4.5,
+      "learning_rate": 4.4483361492740184e-07,
+      "loss": 1.834,
+      "step": 658
+    },
+    {
+      "epoch": 0.8726165254237288,
+      "grad_norm": 4.34375,
+      "learning_rate": 4.3584402339366174e-07,
+      "loss": 2.0094,
+      "step": 659
+    },
+    {
+      "epoch": 0.8739406779661016,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.2694205548510395e-07,
+      "loss": 2.0378,
+      "step": 660
+    },
+    {
+      "epoch": 0.8752648305084746,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.1812788210357836e-07,
+      "loss": 2.1089,
+      "step": 661
+    },
+    {
+      "epoch": 0.8765889830508474,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.0940167246543595e-07,
+      "loss": 1.8608,
+      "step": 662
+    },
+    {
+      "epoch": 0.8779131355932204,
+      "grad_norm": 3.75,
+      "learning_rate": 4.007635940982857e-07,
+      "loss": 2.1243,
+      "step": 663
+    },
+    {
+      "epoch": 0.8792372881355932,
+      "grad_norm": 3.921875,
+      "learning_rate": 3.9221381283776725e-07,
+      "loss": 1.9281,
+      "step": 664
+    },
+    {
+      "epoch": 0.8805614406779662,
+      "grad_norm": 3.71875,
+      "learning_rate": 3.8375249282437743e-07,
+      "loss": 1.9645,
+      "step": 665
+    },
+    {
+      "epoch": 0.881885593220339,
+      "grad_norm": 4.34375,
+      "learning_rate": 3.7537979650031755e-07,
+      "loss": 1.8655,
+      "step": 666
+    },
+    {
+      "epoch": 0.8832097457627118,
+      "grad_norm": 3.65625,
+      "learning_rate": 3.670958846063677e-07,
+      "loss": 2.1663,
+      "step": 667
+    },
+    {
+      "epoch": 0.8845338983050848,
+      "grad_norm": 3.703125,
+      "learning_rate": 3.589009161788104e-07,
+      "loss": 1.7185,
+      "step": 668
+    },
+    {
+      "epoch": 0.8858580508474576,
+      "grad_norm": 3.71875,
+      "learning_rate": 3.507950485463707e-07,
+      "loss": 1.9961,
+      "step": 669
+    },
+    {
+      "epoch": 0.8871822033898306,
+      "grad_norm": 3.71875,
+      "learning_rate": 3.427784373271975e-07,
+      "loss": 1.9487,
+      "step": 670
+    },
+    {
+      "epoch": 0.8885063559322034,
+      "grad_norm": 3.765625,
+      "learning_rate": 3.3485123642587657e-07,
+      "loss": 1.9226,
+      "step": 671
+    },
+    {
+      "epoch": 0.8898305084745762,
+      "grad_norm": 3.84375,
+      "learning_rate": 3.270135980304784e-07,
+      "loss": 2.0407,
+      "step": 672
+    },
+    {
+      "epoch": 0.8911546610169492,
+      "grad_norm": 3.703125,
+      "learning_rate": 3.192656726096277e-07,
+      "loss": 1.8786,
+      "step": 673
+    },
+    {
+      "epoch": 0.892478813559322,
+      "grad_norm": 3.859375,
+      "learning_rate": 3.116076089096265e-07,
+      "loss": 2.0147,
+      "step": 674
+    },
+    {
+      "epoch": 0.893802966101695,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.0403955395159057e-07,
+      "loss": 1.9393,
+      "step": 675
+    },
+    {
+      "epoch": 0.8951271186440678,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.965616530286275e-07,
+      "loss": 1.9254,
+      "step": 676
+    },
+    {
+      "epoch": 0.8964512711864406,
+      "grad_norm": 4.5,
+      "learning_rate": 2.8917404970305096e-07,
+      "loss": 2.1829,
+      "step": 677
+    },
+    {
+      "epoch": 0.8977754237288136,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.818768858036208e-07,
+      "loss": 1.8268,
+      "step": 678
+    },
+    {
+      "epoch": 0.8990995762711864,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.74670301422823e-07,
+      "loss": 2.084,
+      "step": 679
+    },
+    {
+      "epoch": 0.9004237288135594,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.6755443491417786e-07,
+      "loss": 2.0448,
+      "step": 680
+    },
+    {
+      "epoch": 0.9017478813559322,
+      "grad_norm": 3.75,
+      "learning_rate": 2.605294228895855e-07,
+      "loss": 2.0041,
+      "step": 681
+    },
+    {
+      "epoch": 0.903072033898305,
+      "grad_norm": 3.625,
+      "learning_rate": 2.5359540021670124e-07,
+      "loss": 2.0403,
+      "step": 682
+    },
+    {
+      "epoch": 0.904396186440678,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.467525000163523e-07,
+      "loss": 2.0069,
+      "step": 683
+    },
+    {
+      "epoch": 0.9057203389830508,
+      "grad_norm": 3.8125,
+      "learning_rate": 2.4000085365997017e-07,
+      "loss": 1.9129,
+      "step": 684
+    },
+    {
+      "epoch": 0.9070444915254238,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.3334059076708216e-07,
+      "loss": 1.9656,
+      "step": 685
+    },
+    {
+      "epoch": 0.9083686440677966,
+      "grad_norm": 4.125,
+      "learning_rate": 2.2677183920281342e-07,
+      "loss": 2.111,
+      "step": 686
+    },
+    {
+      "epoch": 0.9096927966101694,
+      "grad_norm": 4.0,
+      "learning_rate": 2.2029472507543727e-07,
+      "loss": 2.005,
+      "step": 687
+    },
+    {
+      "epoch": 0.9110169491525424,
+      "grad_norm": 3.875,
+      "learning_rate": 2.139093727339503e-07,
+      "loss": 2.1353,
+      "step": 688
+    },
+    {
+      "epoch": 0.9123411016949152,
+      "grad_norm": 4.0625,
+      "learning_rate": 2.0761590476568893e-07,
+      "loss": 1.9923,
+      "step": 689
+    },
+    {
+      "epoch": 0.9136652542372882,
+      "grad_norm": 3.96875,
+      "learning_rate": 2.0141444199397177e-07,
+      "loss": 1.9819,
+      "step": 690
+    },
+    {
+      "epoch": 0.914989406779661,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.9530510347578634e-07,
+      "loss": 1.9541,
+      "step": 691
+    },
+    {
+      "epoch": 0.9163135593220338,
+      "grad_norm": 3.625,
+      "learning_rate": 1.892880064994934e-07,
+      "loss": 1.9709,
+      "step": 692
+    },
+    {
+      "epoch": 0.9176377118644068,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.8336326658258797e-07,
+      "loss": 2.0905,
+      "step": 693
+    },
+    {
+      "epoch": 0.9189618644067796,
+      "grad_norm": 3.953125,
+      "learning_rate": 1.7753099746947e-07,
+      "loss": 2.0333,
+      "step": 694
+    },
+    {
+      "epoch": 0.9202860169491526,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.7179131112926628e-07,
+      "loss": 1.8526,
+      "step": 695
+    },
+    {
+      "epoch": 0.9216101694915254,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.66144317753682e-07,
+      "loss": 2.1094,
+      "step": 696
+    },
+    {
+      "epoch": 0.9229343220338984,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.6059012575488197e-07,
+      "loss": 2.1094,
+      "step": 697
+    },
+    {
+      "epoch": 0.9242584745762712,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.551288417634106e-07,
+      "loss": 2.2161,
+      "step": 698
+    },
+    {
+      "epoch": 0.925582627118644,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.497605706261457e-07,
+      "loss": 2.0676,
+      "step": 699
+    },
+    {
+      "epoch": 0.926906779661017,
+      "grad_norm": 3.984375,
+      "learning_rate": 1.4448541540428351e-07,
+      "loss": 2.0977,
+      "step": 700
+    },
+    {
+      "epoch": 0.9282309322033898,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.3930347737136195e-07,
+      "loss": 2.1889,
+      "step": 701
+    },
+    {
+      "epoch": 0.9295550847457628,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.3421485601131768e-07,
+      "loss": 1.9644,
+      "step": 702
+    },
+    {
+      "epoch": 0.9308792372881356,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.292196490165698e-07,
+      "loss": 2.0876,
+      "step": 703
+    },
+    {
+      "epoch": 0.9322033898305084,
+      "grad_norm": 6.75,
+      "learning_rate": 1.2431795228615372e-07,
+      "loss": 1.8822,
+      "step": 704
+    },
+    {
+      "epoch": 0.9335275423728814,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.1950985992387197e-07,
+      "loss": 2.1327,
+      "step": 705
+    },
+    {
+      "epoch": 0.9348516949152542,
+      "grad_norm": 3.75,
+      "learning_rate": 1.1479546423649179e-07,
+      "loss": 2.0094,
+      "step": 706
+    },
+    {
+      "epoch": 0.9361758474576272,
+      "grad_norm": 3.875,
+      "learning_rate": 1.1017485573197151e-07,
+      "loss": 1.8658,
+      "step": 707
+    },
+    {
+      "epoch": 0.9375,
+      "grad_norm": 3.625,
+      "learning_rate": 1.0564812311772422e-07,
+      "loss": 1.9581,
+      "step": 708
+    },
+    {
+      "epoch": 0.9388241525423728,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.0121535329891353e-07,
+      "loss": 2.1984,
+      "step": 709
+    },
+    {
+      "epoch": 0.9401483050847458,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.687663137678605e-08,
+      "loss": 1.9161,
+      "step": 710
+    },
+    {
+      "epoch": 0.9414724576271186,
+      "grad_norm": 3.71875,
+      "learning_rate": 9.263204064703546e-08,
+      "loss": 1.9396,
+      "step": 711
+    },
+    {
+      "epoch": 0.9427966101694916,
+      "grad_norm": 4.125,
+      "learning_rate": 8.848166259820657e-08,
+      "loss": 1.9592,
+      "step": 712
+    },
+    {
+      "epoch": 0.9441207627118644,
+      "grad_norm": 3.875,
+      "learning_rate": 8.442557691013042e-08,
+      "loss": 2.1242,
+      "step": 713
+    },
+    {
+      "epoch": 0.9454449152542372,
+      "grad_norm": 3.9375,
+      "learning_rate": 8.046386145239005e-08,
+      "loss": 2.0307,
+      "step": 714
+    },
+    {
+      "epoch": 0.9467690677966102,
+      "grad_norm": 3.828125,
+      "learning_rate": 7.659659228283212e-08,
+      "loss": 2.2161,
+      "step": 715
+    },
+    {
+      "epoch": 0.948093220338983,
+      "grad_norm": 3.734375,
+      "learning_rate": 7.282384364610207e-08,
+      "loss": 2.1596,
+      "step": 716
+    },
+    {
+      "epoch": 0.949417372881356,
+      "grad_norm": 3.9375,
+      "learning_rate": 6.914568797222077e-08,
+      "loss": 2.0263,
+      "step": 717
+    },
+    {
+      "epoch": 0.9507415254237288,
+      "grad_norm": 3.765625,
+      "learning_rate": 6.556219587519397e-08,
+      "loss": 1.9829,
+      "step": 718
+    },
+    {
+      "epoch": 0.9520656779661016,
+      "grad_norm": 3.984375,
+      "learning_rate": 6.207343615165562e-08,
+      "loss": 1.9644,
+      "step": 719
+    },
+    {
+      "epoch": 0.9533898305084746,
+      "grad_norm": 3.65625,
+      "learning_rate": 5.867947577954669e-08,
+      "loss": 1.8524,
+      "step": 720
+    },
+    {
+      "epoch": 0.9547139830508474,
+      "grad_norm": 3.71875,
+      "learning_rate": 5.538037991683287e-08,
+      "loss": 1.8625,
+      "step": 721
+    },
+    {
+      "epoch": 0.9560381355932204,
+      "grad_norm": 3.640625,
+      "learning_rate": 5.21762119002478e-08,
+      "loss": 2.1267,
+      "step": 722
+    },
+    {
+      "epoch": 0.9573622881355932,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.906703324408402e-08,
+      "loss": 2.0322,
+      "step": 723
+    },
+    {
+      "epoch": 0.9586864406779662,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.605290363900505e-08,
+      "loss": 1.8449,
+      "step": 724
+    },
+    {
+      "epoch": 0.960010593220339,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.31338809509052e-08,
+      "loss": 2.0254,
+      "step": 725
+    },
+    {
+      "epoch": 0.9613347457627118,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.031002121979433e-08,
+      "loss": 2.0772,
+      "step": 726
+    },
+    {
+      "epoch": 0.9626588983050848,
+      "grad_norm": 4.09375,
+      "learning_rate": 3.758137865872535e-08,
+      "loss": 2.0363,
+      "step": 727
+    },
+    {
+      "epoch": 0.9639830508474576,
+      "grad_norm": 3.875,
+      "learning_rate": 3.494800565275125e-08,
+      "loss": 2.1279,
+      "step": 728
+    },
+    {
+      "epoch": 0.9653072033898306,
+      "grad_norm": 3.734375,
+      "learning_rate": 3.2409952757919696e-08,
+      "loss": 2.011,
+      "step": 729
+    },
+    {
+      "epoch": 0.9666313559322034,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.9967268700303864e-08,
+      "loss": 1.9656,
+      "step": 730
+    },
+    {
+      "epoch": 0.9679555084745762,
+      "grad_norm": 3.84375,
+      "learning_rate": 2.7620000375064848e-08,
+      "loss": 1.9544,
+      "step": 731
+    },
+    {
+      "epoch": 0.9692796610169492,
+      "grad_norm": 3.875,
+      "learning_rate": 2.5368192845553473e-08,
+      "loss": 2.0779,
+      "step": 732
+    },
+    {
+      "epoch": 0.970603813559322,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.321188934244323e-08,
+      "loss": 2.115,
+      "step": 733
+    },
+    {
+      "epoch": 0.971927966101695,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.115113126290258e-08,
+      "loss": 2.0346,
+      "step": 734
+    },
+    {
+      "epoch": 0.9732521186440678,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.918595816979729e-08,
+      "loss": 1.9826,
+      "step": 735
+    },
+    {
+      "epoch": 0.9745762711864406,
+      "grad_norm": 3.953125,
+      "learning_rate": 1.7316407790933242e-08,
+      "loss": 1.9392,
+      "step": 736
+    },
+    {
+      "epoch": 0.9759004237288136,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.554251601833201e-08,
+      "loss": 1.8802,
+      "step": 737
+    },
+    {
+      "epoch": 0.9772245762711864,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.3864316907539754e-08,
+      "loss": 2.041,
+      "step": 738
+    },
+    {
+      "epoch": 0.9785487288135594,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.2281842676974964e-08,
+      "loss": 2.0679,
+      "step": 739
+    },
+    {
+      "epoch": 0.9798728813559322,
+      "grad_norm": 3.875,
+      "learning_rate": 1.0795123707312283e-08,
+      "loss": 2.0577,
+      "step": 740
+    },
+    {
+      "epoch": 0.981197033898305,
+      "grad_norm": 3.890625,
+      "learning_rate": 9.404188540892977e-09,
+      "loss": 2.0322,
+      "step": 741
+    },
+    {
+      "epoch": 0.982521186440678,
+      "grad_norm": 3.65625,
+      "learning_rate": 8.1090638811826e-09,
+      "loss": 2.0473,
+      "step": 742
+    },
+    {
+      "epoch": 0.9838453389830508,
+      "grad_norm": 3.921875,
+      "learning_rate": 6.9097745922580564e-09,
+      "loss": 2.1947,
+      "step": 743
+    },
+    {
+      "epoch": 0.9851694915254238,
+      "grad_norm": 3.859375,
+      "learning_rate": 5.806343698325223e-09,
+      "loss": 2.0407,
+      "step": 744
+    },
+    {
+      "epoch": 0.9864936440677966,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.79879238328318e-09,
+      "loss": 2.0936,
+      "step": 745
+    },
+    {
+      "epoch": 0.9878177966101694,
+      "grad_norm": 3.6875,
+      "learning_rate": 3.887139990313427e-09,
+      "loss": 2.0415,
+      "step": 746
+    },
+    {
+      "epoch": 0.9891419491525424,
+      "grad_norm": 3.71875,
+      "learning_rate": 3.0714040215101825e-09,
+      "loss": 1.9758,
+      "step": 747
+    },
+    {
+      "epoch": 0.9904661016949152,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.3516001375439856e-09,
+      "loss": 1.9206,
+      "step": 748
+    },
+    {
+      "epoch": 0.9917902542372882,
+      "grad_norm": 3.75,
+      "learning_rate": 1.7277421573608234e-09,
+      "loss": 2.1105,
+      "step": 749
+    },
+    {
+      "epoch": 0.993114406779661,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.1998420579173442e-09,
+      "loss": 1.8936,
+      "step": 750
+    },
+    {
+      "epoch": 0.9944385593220338,
+      "grad_norm": 4.0625,
+      "learning_rate": 7.67909973950487e-10,
+      "loss": 2.0214,
+      "step": 751
+    },
+    {
+      "epoch": 0.9957627118644068,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.3195419778319095e-10,
+      "loss": 2.0709,
+      "step": 752
+    },
+    {
+      "epoch": 0.9970868644067796,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.919811791650794e-10,
+      "loss": 2.0086,
+      "step": 753
+    },
+    {
+      "epoch": 0.9984110169491526,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.7995525148114564e-11,
+      "loss": 2.0926,
+      "step": 754
+    },
+    {
+      "epoch": 0.9997351694915254,
+      "grad_norm": 3.875,
+      "learning_rate": 0.0,
+      "loss": 2.1509,
+      "step": 755
+    },
+    {
+      "epoch": 0.9997351694915254,
+      "step": 755,
+      "total_flos": 1.2357164232319795e+18,
+      "train_loss": 2.099150818073197,
+      "train_runtime": 25557.7164,
+      "train_samples_per_second": 0.591,
+      "train_steps_per_second": 0.03
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 755,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "total_flos": 1.2357164232319795e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}