End of training

Browse files

Files changed (6) hide show

README.md +11 -1
all_results.json +12 -12
eval_results.json +8 -8
runs/Oct31_18-38-12_gtown-28NZK54/events.out.tfevents.1730405768.gtown-28NZK54.1910.1 +3 -0
train_results.json +7 -7
trainer_state.json +1027 -1027

README.md CHANGED Viewed

@@ -3,7 +3,11 @@ library_name: transformers
 license: apache-2.0
 base_model: google/vit-base-patch16-224-in21k
 tags:
 - generated_from_trainer
 model-index:
 - name: vit-base-patch16-224-in21k
   results: []
@@ -14,7 +18,13 @@ should probably proofread and complete it, then remove this comment. -->
 # vit-base-patch16-224-in21k
-This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on an unknown dataset.
 ## Model description

 license: apache-2.0
 base_model: google/vit-base-patch16-224-in21k
 tags:
+- image-classification
+- vision
 - generated_from_trainer
+metrics:
+- accuracy
 model-index:
 - name: vit-base-patch16-224-in21k
   results: []
 # vit-base-patch16-224-in21k
+This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the chainyo/rvl-cdip dataset.
+It achieves the following results on the evaluation set:
+- Loss: nan
+- Accuracy: 0.0630
+- Memory Allocated (gb): 1.49
+- Max Memory Allocated (gb): 2.1
+- Total Memory Available (gb): 126.62
 ## Model description

all_results.json CHANGED Viewed

@@ -1,16 +1,16 @@
 {
     "epoch": 3.0,
-    "eval_accuracy": 0.8788958333333333,
-    "eval_loss": 0.4218454658985138,
-    "eval_runtime": 360.3065,
-    "eval_samples_per_second": 133.22,
-    "eval_steps_per_second": 2.082,
-    "max_memory_allocated (GB)": 2.09,
-    "memory_allocated (GB)": 1.47,
     "total_flos": 6.324139790696448e+19,
-    "total_memory_available (GB)": 94.62,
-    "train_loss": 0.8091567421707453,
-    "train_runtime": 5702.0513,
-    "train_samples_per_second": 143.106,
-    "train_steps_per_second": 17.888
 }

 {
     "epoch": 3.0,
+    "eval_accuracy": 0.06302083333333333,
+    "eval_loss": NaN,
+    "eval_runtime": 244.7131,
+    "eval_samples_per_second": 196.148,
+    "eval_steps_per_second": 3.065,
+    "max_memory_allocated (GB)": 2.1,
+    "memory_allocated (GB)": 1.49,
     "total_flos": 6.324139790696448e+19,
+    "total_memory_available (GB)": 126.62,
+    "train_loss": NaN,
+    "train_runtime": 4358.6451,
+    "train_samples_per_second": 187.213,
+    "train_steps_per_second": 23.402
 }

eval_results.json CHANGED Viewed

@@ -1,11 +1,11 @@
 {
     "epoch": 3.0,
-    "eval_accuracy": 0.8788958333333333,
-    "eval_loss": 0.4218454658985138,
-    "eval_runtime": 360.3065,
-    "eval_samples_per_second": 133.22,
-    "eval_steps_per_second": 2.082,
-    "max_memory_allocated (GB)": 2.09,
-    "memory_allocated (GB)": 1.47,
-    "total_memory_available (GB)": 94.62
 }

 {
     "epoch": 3.0,
+    "eval_accuracy": 0.06302083333333333,
+    "eval_loss": NaN,
+    "eval_runtime": 244.7131,
+    "eval_samples_per_second": 196.148,
+    "eval_steps_per_second": 3.065,
+    "max_memory_allocated (GB)": 2.1,
+    "memory_allocated (GB)": 1.49,
+    "total_memory_available (GB)": 126.62
 }

runs/Oct31_18-38-12_gtown-28NZK54/events.out.tfevents.1730405768.gtown-28NZK54.1910.1 ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:ac7efa4c17143c0b19ccc8d2c087b372bb671be3690a5427f79e85a0471fae57
+size 580

train_results.json CHANGED Viewed

@@ -1,11 +1,11 @@
 {
     "epoch": 3.0,
-    "max_memory_allocated (GB)": 2.09,
-    "memory_allocated (GB)": 1.44,
     "total_flos": 6.324139790696448e+19,
-    "total_memory_available (GB)": 94.62,
-    "train_loss": 0.8091567421707453,
-    "train_runtime": 5702.0513,
-    "train_samples_per_second": 143.106,
-    "train_steps_per_second": 17.888
 }

 {
     "epoch": 3.0,
+    "max_memory_allocated (GB)": 2.1,
+    "memory_allocated (GB)": 1.46,
     "total_flos": 6.324139790696448e+19,
+    "total_memory_available (GB)": 126.62,
+    "train_loss": NaN,
+    "train_runtime": 4358.6451,
+    "train_samples_per_second": 187.213,
+    "train_steps_per_second": 23.402
 }

trainer_state.json CHANGED Viewed

@@ -10,2055 +10,2055 @@
   "log_history": [
     {
       "epoch": 0.014705882352941176,
-      "grad_norm": 2.69562029838562,
       "learning_rate": 4.975490196078432e-05,
-      "loss": 2.0405,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.029411764705882353,
-      "grad_norm": 4.156198978424072,
       "learning_rate": 4.9509803921568634e-05,
-      "loss": 1.6006,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 1000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.04411764705882353,
-      "grad_norm": 4.104771137237549,
       "learning_rate": 4.9264705882352944e-05,
-      "loss": 1.4982,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 1500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.058823529411764705,
-      "grad_norm": 5.6248674392700195,
       "learning_rate": 4.901960784313725e-05,
-      "loss": 1.3869,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 2000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.07352941176470588,
-      "grad_norm": 3.814528226852417,
       "learning_rate": 4.877450980392157e-05,
-      "loss": 1.361,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 2500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.08823529411764706,
-      "grad_norm": 4.103758811950684,
       "learning_rate": 4.8529411764705885e-05,
-      "loss": 1.2968,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 3000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.10294117647058823,
-      "grad_norm": 4.155980587005615,
       "learning_rate": 4.82843137254902e-05,
-      "loss": 1.2537,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 3500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.11764705882352941,
-      "grad_norm": 5.898636341094971,
       "learning_rate": 4.803921568627452e-05,
-      "loss": 1.201,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 4000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.1323529411764706,
-      "grad_norm": 4.946361541748047,
       "learning_rate": 4.7794117647058826e-05,
-      "loss": 1.244,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 4500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.14705882352941177,
-      "grad_norm": 3.6475930213928223,
       "learning_rate": 4.7549019607843135e-05,
-      "loss": 1.2188,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 5000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.16176470588235295,
-      "grad_norm": 6.001659870147705,
       "learning_rate": 4.730392156862745e-05,
-      "loss": 1.1643,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 5500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.17647058823529413,
-      "grad_norm": 4.4559526443481445,
       "learning_rate": 4.705882352941177e-05,
-      "loss": 1.1959,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 6000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.19117647058823528,
-      "grad_norm": 1.5984530448913574,
       "learning_rate": 4.681372549019608e-05,
-      "loss": 1.1349,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 6500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.20588235294117646,
-      "grad_norm": 3.468794345855713,
       "learning_rate": 4.656862745098039e-05,
-      "loss": 1.1358,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 7000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.22058823529411764,
-      "grad_norm": 5.2478485107421875,
       "learning_rate": 4.632352941176471e-05,
-      "loss": 1.0712,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 7500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.23529411764705882,
-      "grad_norm": 4.6103739738464355,
       "learning_rate": 4.607843137254902e-05,
-      "loss": 1.089,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 8000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.25,
-      "grad_norm": 6.615376949310303,
       "learning_rate": 4.5833333333333334e-05,
-      "loss": 1.0837,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 8500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.2647058823529412,
-      "grad_norm": 8.597518920898438,
       "learning_rate": 4.558823529411765e-05,
-      "loss": 1.0957,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 9000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.27941176470588236,
-      "grad_norm": 4.671794891357422,
       "learning_rate": 4.5343137254901966e-05,
-      "loss": 1.0809,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 9500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.29411764705882354,
-      "grad_norm": 6.989711761474609,
       "learning_rate": 4.5098039215686275e-05,
-      "loss": 1.0722,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 10000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.3088235294117647,
-      "grad_norm": 4.134768486022949,
       "learning_rate": 4.485294117647059e-05,
-      "loss": 1.06,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 10500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.3235294117647059,
-      "grad_norm": 7.303393840789795,
       "learning_rate": 4.460784313725491e-05,
-      "loss": 0.9973,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 11000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.3382352941176471,
-      "grad_norm": 5.387753486633301,
       "learning_rate": 4.4362745098039216e-05,
-      "loss": 1.0572,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 11500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.35294117647058826,
-      "grad_norm": 4.046933650970459,
       "learning_rate": 4.411764705882353e-05,
-      "loss": 1.0197,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 12000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.36764705882352944,
-      "grad_norm": 7.536805152893066,
       "learning_rate": 4.387254901960784e-05,
-      "loss": 1.0116,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 12500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.38235294117647056,
-      "grad_norm": 7.579131126403809,
       "learning_rate": 4.362745098039216e-05,
-      "loss": 1.0145,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 13000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.39705882352941174,
-      "grad_norm": 7.731176376342773,
       "learning_rate": 4.3382352941176474e-05,
-      "loss": 1.016,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 13500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.4117647058823529,
-      "grad_norm": 7.9315690994262695,
       "learning_rate": 4.313725490196079e-05,
-      "loss": 0.9855,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 14000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.4264705882352941,
-      "grad_norm": 3.403750419616699,
       "learning_rate": 4.28921568627451e-05,
-      "loss": 0.969,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 14500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.4411764705882353,
-      "grad_norm": 3.1541624069213867,
       "learning_rate": 4.2647058823529415e-05,
-      "loss": 0.9733,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 15000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.45588235294117646,
-      "grad_norm": 6.017275333404541,
       "learning_rate": 4.2401960784313724e-05,
-      "loss": 0.9791,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 15500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.47058823529411764,
-      "grad_norm": 5.537393569946289,
       "learning_rate": 4.215686274509804e-05,
-      "loss": 0.9449,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 16000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.4852941176470588,
-      "grad_norm": 6.600454330444336,
       "learning_rate": 4.1911764705882356e-05,
-      "loss": 0.9643,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 16500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.5,
-      "grad_norm": 5.120028972625732,
       "learning_rate": 4.166666666666667e-05,
-      "loss": 0.9715,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 17000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.5147058823529411,
-      "grad_norm": 3.404690742492676,
       "learning_rate": 4.142156862745099e-05,
-      "loss": 0.9885,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 17500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.5294117647058824,
-      "grad_norm": 2.9108402729034424,
       "learning_rate": 4.11764705882353e-05,
-      "loss": 0.9777,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 18000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.5441176470588235,
-      "grad_norm": 6.1733832359313965,
       "learning_rate": 4.0931372549019607e-05,
-      "loss": 0.9525,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 18500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.5588235294117647,
-      "grad_norm": 2.5774946212768555,
       "learning_rate": 4.068627450980392e-05,
-      "loss": 0.9244,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 19000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.5735294117647058,
-      "grad_norm": 2.9848501682281494,
       "learning_rate": 4.044117647058824e-05,
-      "loss": 0.9585,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 19500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.5882352941176471,
-      "grad_norm": 3.7758054733276367,
       "learning_rate": 4.0196078431372555e-05,
-      "loss": 0.9449,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 20000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.6029411764705882,
-      "grad_norm": 7.037225246429443,
       "learning_rate": 3.9950980392156864e-05,
-      "loss": 0.9551,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 20500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.6176470588235294,
-      "grad_norm": 2.501617670059204,
       "learning_rate": 3.970588235294117e-05,
-      "loss": 0.9396,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 21000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.6323529411764706,
-      "grad_norm": 3.3458075523376465,
       "learning_rate": 3.946078431372549e-05,
-      "loss": 0.9053,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 21500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.6470588235294118,
-      "grad_norm": 6.687346458435059,
       "learning_rate": 3.9215686274509805e-05,
-      "loss": 0.9441,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 22000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.6617647058823529,
-      "grad_norm": 4.228150844573975,
       "learning_rate": 3.897058823529412e-05,
-      "loss": 0.9013,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 22500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.6764705882352942,
-      "grad_norm": 5.184825420379639,
       "learning_rate": 3.872549019607844e-05,
-      "loss": 0.8978,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 23000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.6911764705882353,
-      "grad_norm": 4.4502058029174805,
       "learning_rate": 3.8480392156862746e-05,
-      "loss": 0.9099,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 23500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.7058823529411765,
-      "grad_norm": 3.760995388031006,
       "learning_rate": 3.8235294117647055e-05,
-      "loss": 0.9156,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 24000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.7205882352941176,
-      "grad_norm": 8.011835098266602,
       "learning_rate": 3.799019607843137e-05,
-      "loss": 0.8686,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 24500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.7352941176470589,
-      "grad_norm": 6.8909382820129395,
       "learning_rate": 3.774509803921569e-05,
-      "loss": 0.8945,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 25000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.75,
-      "grad_norm": 7.357186317443848,
       "learning_rate": 3.7500000000000003e-05,
-      "loss": 0.8787,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 25500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.7647058823529411,
-      "grad_norm": 7.247834205627441,
       "learning_rate": 3.725490196078432e-05,
-      "loss": 0.8591,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 26000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.7794117647058824,
-      "grad_norm": 0.40234723687171936,
       "learning_rate": 3.700980392156863e-05,
-      "loss": 0.8748,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 26500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.7941176470588235,
-      "grad_norm": 8.666840553283691,
       "learning_rate": 3.6764705882352945e-05,
-      "loss": 0.8754,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 27000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.8088235294117647,
-      "grad_norm": 5.52506160736084,
       "learning_rate": 3.6519607843137254e-05,
-      "loss": 0.8865,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 27500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.8235294117647058,
-      "grad_norm": 4.968148231506348,
       "learning_rate": 3.627450980392157e-05,
-      "loss": 0.8651,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 28000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.8382352941176471,
-      "grad_norm": 2.7325375080108643,
       "learning_rate": 3.6029411764705886e-05,
-      "loss": 0.8577,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 28500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.8529411764705882,
-      "grad_norm": 2.859745502471924,
       "learning_rate": 3.5784313725490195e-05,
-      "loss": 0.8683,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 29000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.8676470588235294,
-      "grad_norm": 4.353167533874512,
       "learning_rate": 3.553921568627451e-05,
-      "loss": 0.8804,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 29500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.8823529411764706,
-      "grad_norm": 8.247920989990234,
       "learning_rate": 3.529411764705883e-05,
-      "loss": 0.8303,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 30000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.8970588235294118,
-      "grad_norm": 6.413235664367676,
       "learning_rate": 3.5049019607843136e-05,
-      "loss": 0.8827,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 30500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.9117647058823529,
-      "grad_norm": 6.1504621505737305,
       "learning_rate": 3.480392156862745e-05,
-      "loss": 0.8411,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 31000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.9264705882352942,
-      "grad_norm": 3.2474753856658936,
       "learning_rate": 3.455882352941177e-05,
-      "loss": 0.8524,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 31500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.9411764705882353,
-      "grad_norm": 4.47405481338501,
       "learning_rate": 3.431372549019608e-05,
-      "loss": 0.8539,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 32000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.9558823529411765,
-      "grad_norm": 14.277088165283203,
       "learning_rate": 3.4068627450980394e-05,
-      "loss": 0.8531,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 32500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.9705882352941176,
-      "grad_norm": 0.5225111842155457,
       "learning_rate": 3.382352941176471e-05,
-      "loss": 0.8423,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 33000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 0.9852941176470589,
-      "grad_norm": 5.679318904876709,
       "learning_rate": 3.357843137254902e-05,
-      "loss": 0.8335,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 33500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.0,
-      "grad_norm": 4.732542514801025,
       "learning_rate": 3.3333333333333335e-05,
-      "loss": 0.8728,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 34000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.0147058823529411,
-      "grad_norm": 5.163674354553223,
       "learning_rate": 3.308823529411765e-05,
-      "loss": 0.8364,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 34500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.0294117647058822,
-      "grad_norm": 4.340713024139404,
       "learning_rate": 3.284313725490196e-05,
-      "loss": 0.795,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 35000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.0441176470588236,
-      "grad_norm": 1.8152037858963013,
       "learning_rate": 3.2598039215686276e-05,
-      "loss": 0.8103,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 35500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.0588235294117647,
-      "grad_norm": 4.854606628417969,
       "learning_rate": 3.235294117647059e-05,
-      "loss": 0.838,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 36000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.0735294117647058,
-      "grad_norm": 9.439714431762695,
       "learning_rate": 3.210784313725491e-05,
-      "loss": 0.82,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 36500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.088235294117647,
-      "grad_norm": 4.896808624267578,
       "learning_rate": 3.186274509803922e-05,
-      "loss": 0.854,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 37000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.1029411764705883,
-      "grad_norm": 5.517809867858887,
       "learning_rate": 3.161764705882353e-05,
-      "loss": 0.8085,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 37500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.1176470588235294,
-      "grad_norm": 4.532939910888672,
       "learning_rate": 3.137254901960784e-05,
-      "loss": 0.7907,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 38000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.1323529411764706,
-      "grad_norm": 2.4086740016937256,
       "learning_rate": 3.112745098039216e-05,
-      "loss": 0.767,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 38500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.1470588235294117,
-      "grad_norm": 3.4542815685272217,
       "learning_rate": 3.0882352941176475e-05,
-      "loss": 0.8148,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 39000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.161764705882353,
-      "grad_norm": 3.529639959335327,
       "learning_rate": 3.063725490196079e-05,
-      "loss": 0.7987,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 39500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.1764705882352942,
-      "grad_norm": 4.117021560668945,
       "learning_rate": 3.0392156862745097e-05,
-      "loss": 0.7714,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 40000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.1911764705882353,
-      "grad_norm": 6.675631523132324,
       "learning_rate": 3.0147058823529413e-05,
-      "loss": 0.7904,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 40500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.2058823529411764,
-      "grad_norm": 10.180689811706543,
       "learning_rate": 2.9901960784313725e-05,
-      "loss": 0.809,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 41000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.2205882352941178,
-      "grad_norm": 4.973780632019043,
       "learning_rate": 2.965686274509804e-05,
-      "loss": 0.8098,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 41500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.2352941176470589,
-      "grad_norm": 2.8854587078094482,
       "learning_rate": 2.9411764705882354e-05,
-      "loss": 0.7711,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 42000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.25,
-      "grad_norm": 3.0195820331573486,
       "learning_rate": 2.916666666666667e-05,
-      "loss": 0.8153,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 42500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.2647058823529411,
-      "grad_norm": 7.375023365020752,
       "learning_rate": 2.8921568627450986e-05,
-      "loss": 0.7773,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 43000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.2794117647058822,
-      "grad_norm": 2.4358959197998047,
       "learning_rate": 2.8676470588235295e-05,
-      "loss": 0.7846,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 43500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.2941176470588236,
-      "grad_norm": 3.8631513118743896,
       "learning_rate": 2.8431372549019608e-05,
-      "loss": 0.7751,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 44000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.3088235294117647,
-      "grad_norm": 4.434110164642334,
       "learning_rate": 2.8186274509803924e-05,
-      "loss": 0.7435,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 44500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.3235294117647058,
-      "grad_norm": 0.8194059133529663,
       "learning_rate": 2.7941176470588236e-05,
-      "loss": 0.7956,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 45000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.3382352941176472,
-      "grad_norm": 13.403727531433105,
       "learning_rate": 2.7696078431372552e-05,
-      "loss": 0.7502,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 45500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.3529411764705883,
-      "grad_norm": 6.71666145324707,
       "learning_rate": 2.7450980392156865e-05,
-      "loss": 0.7996,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 46000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.3676470588235294,
-      "grad_norm": 5.671672344207764,
       "learning_rate": 2.7205882352941174e-05,
-      "loss": 0.785,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 46500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.3823529411764706,
-      "grad_norm": 7.542393207550049,
       "learning_rate": 2.696078431372549e-05,
-      "loss": 0.751,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 47000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.3970588235294117,
-      "grad_norm": 9.15674114227295,
       "learning_rate": 2.6715686274509806e-05,
-      "loss": 0.7809,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 47500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.4117647058823528,
-      "grad_norm": 9.020813941955566,
       "learning_rate": 2.647058823529412e-05,
-      "loss": 0.775,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 48000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.4264705882352942,
-      "grad_norm": 5.843054294586182,
       "learning_rate": 2.6225490196078435e-05,
-      "loss": 0.7526,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 48500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.4411764705882353,
-      "grad_norm": 6.700314521789551,
       "learning_rate": 2.5980392156862747e-05,
-      "loss": 0.7809,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 49000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.4558823529411764,
-      "grad_norm": 4.38136625289917,
       "learning_rate": 2.5735294117647057e-05,
-      "loss": 0.8103,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 49500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.4705882352941178,
-      "grad_norm": 4.148223400115967,
       "learning_rate": 2.5490196078431373e-05,
-      "loss": 0.749,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 50000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.4852941176470589,
-      "grad_norm": 3.625410795211792,
       "learning_rate": 2.5245098039215685e-05,
-      "loss": 0.7584,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 50500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.5,
-      "grad_norm": 4.744174480438232,
       "learning_rate": 2.5e-05,
-      "loss": 0.7611,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 51000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.5147058823529411,
-      "grad_norm": 3.2249341011047363,
       "learning_rate": 2.4754901960784317e-05,
-      "loss": 0.7571,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 51500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.5294117647058822,
-      "grad_norm": 20.33920669555664,
       "learning_rate": 2.4509803921568626e-05,
-      "loss": 0.7638,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 52000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.5441176470588234,
-      "grad_norm": 7.30839729309082,
       "learning_rate": 2.4264705882352942e-05,
-      "loss": 0.7334,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 52500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.5588235294117647,
-      "grad_norm": 7.568194389343262,
       "learning_rate": 2.401960784313726e-05,
-      "loss": 0.7409,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 53000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.5735294117647058,
-      "grad_norm": 5.886373519897461,
       "learning_rate": 2.3774509803921568e-05,
-      "loss": 0.7632,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 53500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.5882352941176472,
-      "grad_norm": 3.079610824584961,
       "learning_rate": 2.3529411764705884e-05,
-      "loss": 0.7436,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 54000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.6029411764705883,
-      "grad_norm": 3.6385908126831055,
       "learning_rate": 2.3284313725490196e-05,
-      "loss": 0.7572,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 54500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.6176470588235294,
-      "grad_norm": 3.804089069366455,
       "learning_rate": 2.303921568627451e-05,
-      "loss": 0.7404,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 55000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.6323529411764706,
-      "grad_norm": 8.9433012008667,
       "learning_rate": 2.2794117647058825e-05,
-      "loss": 0.7431,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 55500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.6470588235294117,
-      "grad_norm": 2.727815628051758,
       "learning_rate": 2.2549019607843138e-05,
-      "loss": 0.7042,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 56000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.6617647058823528,
-      "grad_norm": 10.063603401184082,
       "learning_rate": 2.2303921568627454e-05,
-      "loss": 0.7107,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 56500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.6764705882352942,
-      "grad_norm": 4.510873317718506,
       "learning_rate": 2.2058823529411766e-05,
-      "loss": 0.7535,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 57000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.6911764705882353,
-      "grad_norm": 5.605432510375977,
       "learning_rate": 2.181372549019608e-05,
-      "loss": 0.7505,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 57500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.7058823529411766,
-      "grad_norm": 0.6501856446266174,
       "learning_rate": 2.1568627450980395e-05,
-      "loss": 0.7346,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 58000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.7205882352941178,
-      "grad_norm": 5.737814903259277,
       "learning_rate": 2.1323529411764707e-05,
-      "loss": 0.7569,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 58500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.7352941176470589,
-      "grad_norm": 8.297637939453125,
       "learning_rate": 2.107843137254902e-05,
-      "loss": 0.7421,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 59000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.75,
-      "grad_norm": 4.928978443145752,
       "learning_rate": 2.0833333333333336e-05,
-      "loss": 0.7456,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 59500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.7647058823529411,
-      "grad_norm": 2.3863282203674316,
       "learning_rate": 2.058823529411765e-05,
-      "loss": 0.7474,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 60000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.7794117647058822,
-      "grad_norm": 4.122894287109375,
       "learning_rate": 2.034313725490196e-05,
-      "loss": 0.7338,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 60500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.7941176470588234,
-      "grad_norm": 3.6892104148864746,
       "learning_rate": 2.0098039215686277e-05,
-      "loss": 0.725,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 61000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.8088235294117647,
-      "grad_norm": 6.866763591766357,
       "learning_rate": 1.9852941176470586e-05,
-      "loss": 0.7241,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 61500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.8235294117647058,
-      "grad_norm": 1.1342861652374268,
       "learning_rate": 1.9607843137254903e-05,
-      "loss": 0.7295,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 62000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.8382352941176472,
-      "grad_norm": 2.4428770542144775,
       "learning_rate": 1.936274509803922e-05,
-      "loss": 0.733,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 62500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.8529411764705883,
-      "grad_norm": 9.717529296875,
       "learning_rate": 1.9117647058823528e-05,
-      "loss": 0.7384,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 63000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.8676470588235294,
-      "grad_norm": 2.8925726413726807,
       "learning_rate": 1.8872549019607844e-05,
-      "loss": 0.7347,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 63500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.8823529411764706,
-      "grad_norm": 3.295344352722168,
       "learning_rate": 1.862745098039216e-05,
-      "loss": 0.6938,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 64000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.8970588235294117,
-      "grad_norm": 2.7375340461730957,
       "learning_rate": 1.8382352941176472e-05,
-      "loss": 0.6905,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 64500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.9117647058823528,
-      "grad_norm": 6.166092395782471,
       "learning_rate": 1.8137254901960785e-05,
-      "loss": 0.6869,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 65000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.9264705882352942,
-      "grad_norm": 6.4301958084106445,
       "learning_rate": 1.7892156862745098e-05,
-      "loss": 0.6835,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 65500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.9411764705882353,
-      "grad_norm": 4.680153846740723,
       "learning_rate": 1.7647058823529414e-05,
-      "loss": 0.7159,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 66000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.9558823529411766,
-      "grad_norm": 7.453646183013916,
       "learning_rate": 1.7401960784313726e-05,
-      "loss": 0.7334,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 66500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.9705882352941178,
-      "grad_norm": 5.627362251281738,
       "learning_rate": 1.715686274509804e-05,
-      "loss": 0.6815,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 67000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 1.9852941176470589,
-      "grad_norm": 11.986687660217285,
       "learning_rate": 1.6911764705882355e-05,
-      "loss": 0.6722,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 67500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.0,
-      "grad_norm": 4.421992778778076,
       "learning_rate": 1.6666666666666667e-05,
-      "loss": 0.673,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 68000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.014705882352941,
-      "grad_norm": 7.424468517303467,
       "learning_rate": 1.642156862745098e-05,
-      "loss": 0.6993,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 68500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.0294117647058822,
-      "grad_norm": 5.982107639312744,
       "learning_rate": 1.6176470588235296e-05,
-      "loss": 0.6506,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 69000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.0441176470588234,
-      "grad_norm": 1.392773151397705,
       "learning_rate": 1.593137254901961e-05,
-      "loss": 0.6809,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 69500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.0588235294117645,
-      "grad_norm": 9.153546333312988,
       "learning_rate": 1.568627450980392e-05,
-      "loss": 0.7213,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 70000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.073529411764706,
-      "grad_norm": 7.502923965454102,
       "learning_rate": 1.5441176470588237e-05,
-      "loss": 0.6766,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 70500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.088235294117647,
-      "grad_norm": 5.26950216293335,
       "learning_rate": 1.5196078431372548e-05,
-      "loss": 0.6549,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 71000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.1029411764705883,
-      "grad_norm": 4.774547100067139,
       "learning_rate": 1.4950980392156863e-05,
-      "loss": 0.6979,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 71500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.1176470588235294,
-      "grad_norm": 6.183305740356445,
       "learning_rate": 1.4705882352941177e-05,
-      "loss": 0.6647,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 72000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.1323529411764706,
-      "grad_norm": 6.1788811683654785,
       "learning_rate": 1.4460784313725493e-05,
-      "loss": 0.7104,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 72500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.1470588235294117,
-      "grad_norm": 6.638744831085205,
       "learning_rate": 1.4215686274509804e-05,
-      "loss": 0.6424,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 73000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.161764705882353,
-      "grad_norm": 11.812053680419922,
       "learning_rate": 1.3970588235294118e-05,
-      "loss": 0.7078,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 73500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.176470588235294,
-      "grad_norm": 7.511487007141113,
       "learning_rate": 1.3725490196078432e-05,
-      "loss": 0.6519,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 74000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.1911764705882355,
-      "grad_norm": 4.150969982147217,
       "learning_rate": 1.3480392156862745e-05,
-      "loss": 0.6617,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 74500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.2058823529411766,
-      "grad_norm": 2.6986491680145264,
       "learning_rate": 1.323529411764706e-05,
-      "loss": 0.6712,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 75000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.2205882352941178,
-      "grad_norm": 5.613584041595459,
       "learning_rate": 1.2990196078431374e-05,
-      "loss": 0.6358,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 75500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.235294117647059,
-      "grad_norm": 4.334458351135254,
       "learning_rate": 1.2745098039215686e-05,
-      "loss": 0.7113,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 76000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.25,
-      "grad_norm": 8.442917823791504,
       "learning_rate": 1.25e-05,
-      "loss": 0.6613,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 76500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.264705882352941,
-      "grad_norm": 5.224609851837158,
       "learning_rate": 1.2254901960784313e-05,
-      "loss": 0.6748,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 77000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.2794117647058822,
-      "grad_norm": 2.9958598613739014,
       "learning_rate": 1.200980392156863e-05,
-      "loss": 0.6934,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 77500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.2941176470588234,
-      "grad_norm": 7.674880027770996,
       "learning_rate": 1.1764705882352942e-05,
-      "loss": 0.6294,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 78000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.3088235294117645,
-      "grad_norm": 4.707705497741699,
       "learning_rate": 1.1519607843137254e-05,
-      "loss": 0.6297,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 78500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.323529411764706,
-      "grad_norm": 4.538410663604736,
       "learning_rate": 1.1274509803921569e-05,
-      "loss": 0.6277,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 79000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.338235294117647,
-      "grad_norm": 5.959723949432373,
       "learning_rate": 1.1029411764705883e-05,
-      "loss": 0.6544,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 79500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.3529411764705883,
-      "grad_norm": 3.681849479675293,
       "learning_rate": 1.0784313725490197e-05,
-      "loss": 0.6407,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 80000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.3676470588235294,
-      "grad_norm": 1.0268243551254272,
       "learning_rate": 1.053921568627451e-05,
-      "loss": 0.6442,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 80500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.3823529411764706,
-      "grad_norm": 3.558901786804199,
       "learning_rate": 1.0294117647058824e-05,
-      "loss": 0.6604,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 81000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.3970588235294117,
-      "grad_norm": 4.599427223205566,
       "learning_rate": 1.0049019607843139e-05,
-      "loss": 0.6193,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 81500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.411764705882353,
-      "grad_norm": 3.1800103187561035,
       "learning_rate": 9.803921568627451e-06,
-      "loss": 0.6674,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 82000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.426470588235294,
-      "grad_norm": 3.8055174350738525,
       "learning_rate": 9.558823529411764e-06,
-      "loss": 0.6839,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 82500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.4411764705882355,
-      "grad_norm": 2.8182101249694824,
       "learning_rate": 9.31372549019608e-06,
-      "loss": 0.6537,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 83000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.4558823529411766,
-      "grad_norm": 6.889364719390869,
       "learning_rate": 9.068627450980392e-06,
-      "loss": 0.659,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 83500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.4705882352941178,
-      "grad_norm": 7.853886604309082,
       "learning_rate": 8.823529411764707e-06,
-      "loss": 0.67,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 84000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.485294117647059,
-      "grad_norm": 8.194958686828613,
       "learning_rate": 8.57843137254902e-06,
-      "loss": 0.6585,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 84500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.5,
-      "grad_norm": 4.7262797355651855,
       "learning_rate": 8.333333333333334e-06,
-      "loss": 0.6413,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 85000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.514705882352941,
-      "grad_norm": 7.074121952056885,
       "learning_rate": 8.088235294117648e-06,
-      "loss": 0.6184,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 85500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.5294117647058822,
-      "grad_norm": 5.776811599731445,
       "learning_rate": 7.84313725490196e-06,
-      "loss": 0.6196,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 86000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.5441176470588234,
-      "grad_norm": 1.1076288223266602,
       "learning_rate": 7.598039215686274e-06,
-      "loss": 0.6491,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 86500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.5588235294117645,
-      "grad_norm": 4.6030354499816895,
       "learning_rate": 7.3529411764705884e-06,
-      "loss": 0.668,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 87000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.5735294117647056,
-      "grad_norm": 4.7051682472229,
       "learning_rate": 7.107843137254902e-06,
-      "loss": 0.6404,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 87500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.588235294117647,
-      "grad_norm": 5.557882308959961,
       "learning_rate": 6.862745098039216e-06,
-      "loss": 0.6326,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 88000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.6029411764705883,
-      "grad_norm": 6.156619548797607,
       "learning_rate": 6.61764705882353e-06,
-      "loss": 0.6307,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 88500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.6176470588235294,
-      "grad_norm": 6.322505474090576,
       "learning_rate": 6.372549019607843e-06,
-      "loss": 0.6219,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 89000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.6323529411764706,
-      "grad_norm": 9.32959270477295,
       "learning_rate": 6.127450980392157e-06,
-      "loss": 0.6621,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 89500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.6470588235294117,
-      "grad_norm": 6.305071830749512,
       "learning_rate": 5.882352941176471e-06,
-      "loss": 0.6659,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 90000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.661764705882353,
-      "grad_norm": 6.671431064605713,
       "learning_rate": 5.637254901960784e-06,
-      "loss": 0.6363,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 90500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.6764705882352944,
-      "grad_norm": 7.769582748413086,
       "learning_rate": 5.392156862745099e-06,
-      "loss": 0.5968,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 91000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.6911764705882355,
-      "grad_norm": 5.684257507324219,
       "learning_rate": 5.147058823529412e-06,
-      "loss": 0.6371,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 91500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.7058823529411766,
-      "grad_norm": 2.3605005741119385,
       "learning_rate": 4.901960784313726e-06,
-      "loss": 0.6213,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 92000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.7205882352941178,
-      "grad_norm": 2.693053960800171,
       "learning_rate": 4.65686274509804e-06,
-      "loss": 0.6121,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 92500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.735294117647059,
-      "grad_norm": 5.366748332977295,
       "learning_rate": 4.411764705882353e-06,
-      "loss": 0.6359,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 93000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.75,
-      "grad_norm": 10.606904983520508,
       "learning_rate": 4.166666666666667e-06,
-      "loss": 0.626,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 93500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.764705882352941,
-      "grad_norm": 3.8781898021698,
       "learning_rate": 3.92156862745098e-06,
-      "loss": 0.6125,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 94000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.7794117647058822,
-      "grad_norm": 5.4950480461120605,
       "learning_rate": 3.6764705882352942e-06,
-      "loss": 0.6427,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 94500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.7941176470588234,
-      "grad_norm": 9.753504753112793,
       "learning_rate": 3.431372549019608e-06,
-      "loss": 0.609,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 95000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.8088235294117645,
-      "grad_norm": 8.253406524658203,
       "learning_rate": 3.1862745098039216e-06,
-      "loss": 0.6151,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 95500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.8235294117647056,
-      "grad_norm": 2.7147912979125977,
       "learning_rate": 2.9411764705882355e-06,
-      "loss": 0.6456,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 96000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.838235294117647,
-      "grad_norm": 5.022973537445068,
       "learning_rate": 2.6960784313725493e-06,
-      "loss": 0.6312,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 96500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.8529411764705883,
-      "grad_norm": 12.746963500976562,
       "learning_rate": 2.450980392156863e-06,
-      "loss": 0.6035,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 97000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.8676470588235294,
-      "grad_norm": 4.809776306152344,
       "learning_rate": 2.2058823529411767e-06,
-      "loss": 0.6058,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 97500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.8823529411764706,
-      "grad_norm": 7.464942455291748,
       "learning_rate": 1.96078431372549e-06,
-      "loss": 0.6275,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 98000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.8970588235294117,
-      "grad_norm": 4.378190517425537,
       "learning_rate": 1.715686274509804e-06,
-      "loss": 0.6096,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 98500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.911764705882353,
-      "grad_norm": 6.932071685791016,
       "learning_rate": 1.4705882352941177e-06,
-      "loss": 0.6174,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 99000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.9264705882352944,
-      "grad_norm": 1.794121503829956,
       "learning_rate": 1.2254901960784314e-06,
-      "loss": 0.6046,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 99500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.9411764705882355,
-      "grad_norm": 6.8889055252075195,
       "learning_rate": 9.80392156862745e-07,
-      "loss": 0.6232,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 100000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.9558823529411766,
-      "grad_norm": 5.478596210479736,
       "learning_rate": 7.352941176470589e-07,
-      "loss": 0.6378,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 100500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.9705882352941178,
-      "grad_norm": 6.872597694396973,
       "learning_rate": 4.901960784313725e-07,
-      "loss": 0.6229,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 101000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 2.985294117647059,
-      "grad_norm": 10.850652694702148,
       "learning_rate": 2.4509803921568627e-07,
-      "loss": 0.6352,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 101500,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 3.0,
-      "grad_norm": 4.699875831604004,
       "learning_rate": 0.0,
-      "loss": 0.5992,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 102000,
-      "total_memory_available (GB)": 94.62
     },
     {
       "epoch": 3.0,
-      "max_memory_allocated (GB)": 2.09,
-      "memory_allocated (GB)": 1.44,
       "step": 102000,
       "total_flos": 6.324139790696448e+19,
-      "total_memory_available (GB)": 94.62,
-      "train_loss": 0.8091567421707453,
-      "train_runtime": 5702.0513,
-      "train_samples_per_second": 143.106,
-      "train_steps_per_second": 17.888
     }
   ],
   "logging_steps": 500,

   "log_history": [
     {
       "epoch": 0.014705882352941176,
+      "grad_norm": 2.769134283065796,
       "learning_rate": 4.975490196078432e-05,
+      "loss": 2.0395,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.029411764705882353,
+      "grad_norm": 3.531428813934326,
       "learning_rate": 4.9509803921568634e-05,
+      "loss": 1.5943,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 1000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.04411764705882353,
+      "grad_norm": 7.56320858001709,
       "learning_rate": 4.9264705882352944e-05,
+      "loss": 1.4979,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 1500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.058823529411764705,
+      "grad_norm": 5.256613254547119,
       "learning_rate": 4.901960784313725e-05,
+      "loss": 1.3957,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 2000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.07352941176470588,
+      "grad_norm": 4.642207145690918,
       "learning_rate": 4.877450980392157e-05,
+      "loss": 1.3589,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 2500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.08823529411764706,
+      "grad_norm": 3.366818904876709,
       "learning_rate": 4.8529411764705885e-05,
+      "loss": 1.3006,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 3000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.10294117647058823,
+      "grad_norm": 4.411001205444336,
       "learning_rate": 4.82843137254902e-05,
+      "loss": 1.2582,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 3500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.11764705882352941,
+      "grad_norm": 5.188011646270752,
       "learning_rate": 4.803921568627452e-05,
+      "loss": 1.1948,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 4000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.1323529411764706,
+      "grad_norm": 3.674046754837036,
       "learning_rate": 4.7794117647058826e-05,
+      "loss": 1.2398,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 4500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.14705882352941177,
+      "grad_norm": 6.163153171539307,
       "learning_rate": 4.7549019607843135e-05,
+      "loss": 1.2257,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 5000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.16176470588235295,
+      "grad_norm": 5.589200019836426,
       "learning_rate": 4.730392156862745e-05,
+      "loss": 1.1747,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 5500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.17647058823529413,
+      "grad_norm": 5.379985332489014,
       "learning_rate": 4.705882352941177e-05,
+      "loss": 1.1922,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 6000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.19117647058823528,
+      "grad_norm": 1.5078221559524536,
       "learning_rate": 4.681372549019608e-05,
+      "loss": 1.1355,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 6500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.20588235294117646,
+      "grad_norm": 2.732482433319092,
       "learning_rate": 4.656862745098039e-05,
+      "loss": 1.1359,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 7000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.22058823529411764,
+      "grad_norm": 5.508218765258789,
       "learning_rate": 4.632352941176471e-05,
+      "loss": 1.0726,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 7500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.23529411764705882,
+      "grad_norm": 6.273388385772705,
       "learning_rate": 4.607843137254902e-05,
+      "loss": 1.0777,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 8000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.25,
+      "grad_norm": 7.923452854156494,
       "learning_rate": 4.5833333333333334e-05,
+      "loss": 1.0869,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 8500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.2647058823529412,
+      "grad_norm": 4.915501117706299,
       "learning_rate": 4.558823529411765e-05,
+      "loss": 1.0897,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 9000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.27941176470588236,
+      "grad_norm": 4.274153232574463,
       "learning_rate": 4.5343137254901966e-05,
+      "loss": 1.0816,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 9500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.29411764705882354,
+      "grad_norm": 6.414577960968018,
       "learning_rate": 4.5098039215686275e-05,
+      "loss": 1.0721,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 10000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.3088235294117647,
+      "grad_norm": 5.6046342849731445,
       "learning_rate": 4.485294117647059e-05,
+      "loss": 1.052,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 10500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.3235294117647059,
+      "grad_norm": 6.429250717163086,
       "learning_rate": 4.460784313725491e-05,
+      "loss": 0.9923,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 11000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.3382352941176471,
+      "grad_norm": 3.2989513874053955,
       "learning_rate": 4.4362745098039216e-05,
+      "loss": 1.0657,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 11500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.35294117647058826,
+      "grad_norm": 6.195382118225098,
       "learning_rate": 4.411764705882353e-05,
+      "loss": 1.0001,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 12000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.36764705882352944,
+      "grad_norm": 2.9487133026123047,
       "learning_rate": 4.387254901960784e-05,
+      "loss": 1.0021,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 12500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.38235294117647056,
+      "grad_norm": 7.185413837432861,
       "learning_rate": 4.362745098039216e-05,
+      "loss": 1.0065,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 13000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.39705882352941174,
+      "grad_norm": 5.732925891876221,
       "learning_rate": 4.3382352941176474e-05,
+      "loss": 1.0212,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 13500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.4117647058823529,
+      "grad_norm": 6.557607650756836,
       "learning_rate": 4.313725490196079e-05,
+      "loss": 0.986,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 14000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.4264705882352941,
+      "grad_norm": 4.9897613525390625,
       "learning_rate": 4.28921568627451e-05,
+      "loss": 0.977,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 14500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.4411764705882353,
+      "grad_norm": 2.1532859802246094,
       "learning_rate": 4.2647058823529415e-05,
+      "loss": 0.9729,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 15000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.45588235294117646,
+      "grad_norm": NaN,
       "learning_rate": 4.2401960784313724e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 15500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.47058823529411764,
+      "grad_norm": NaN,
       "learning_rate": 4.215686274509804e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 16000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.4852941176470588,
+      "grad_norm": NaN,
       "learning_rate": 4.1911764705882356e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 16500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.5,
+      "grad_norm": NaN,
       "learning_rate": 4.166666666666667e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 17000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.5147058823529411,
+      "grad_norm": NaN,
       "learning_rate": 4.142156862745099e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 17500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.5294117647058824,
+      "grad_norm": NaN,
       "learning_rate": 4.11764705882353e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 18000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.5441176470588235,
+      "grad_norm": NaN,
       "learning_rate": 4.0931372549019607e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 18500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.5588235294117647,
+      "grad_norm": NaN,
       "learning_rate": 4.068627450980392e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 19000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.5735294117647058,
+      "grad_norm": NaN,
       "learning_rate": 4.044117647058824e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 19500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.5882352941176471,
+      "grad_norm": NaN,
       "learning_rate": 4.0196078431372555e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 20000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.6029411764705882,
+      "grad_norm": NaN,
       "learning_rate": 3.9950980392156864e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 20500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.6176470588235294,
+      "grad_norm": NaN,
       "learning_rate": 3.970588235294117e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 21000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.6323529411764706,
+      "grad_norm": NaN,
       "learning_rate": 3.946078431372549e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 21500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.6470588235294118,
+      "grad_norm": NaN,
       "learning_rate": 3.9215686274509805e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 22000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.6617647058823529,
+      "grad_norm": NaN,
       "learning_rate": 3.897058823529412e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 22500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.6764705882352942,
+      "grad_norm": NaN,
       "learning_rate": 3.872549019607844e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 23000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.6911764705882353,
+      "grad_norm": NaN,
       "learning_rate": 3.8480392156862746e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 23500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.7058823529411765,
+      "grad_norm": NaN,
       "learning_rate": 3.8235294117647055e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 24000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.7205882352941176,
+      "grad_norm": NaN,
       "learning_rate": 3.799019607843137e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 24500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.7352941176470589,
+      "grad_norm": NaN,
       "learning_rate": 3.774509803921569e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 25000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.75,
+      "grad_norm": NaN,
       "learning_rate": 3.7500000000000003e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 25500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.7647058823529411,
+      "grad_norm": NaN,
       "learning_rate": 3.725490196078432e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 26000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.7794117647058824,
+      "grad_norm": NaN,
       "learning_rate": 3.700980392156863e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 26500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.7941176470588235,
+      "grad_norm": NaN,
       "learning_rate": 3.6764705882352945e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 27000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.8088235294117647,
+      "grad_norm": NaN,
       "learning_rate": 3.6519607843137254e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 27500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.8235294117647058,
+      "grad_norm": NaN,
       "learning_rate": 3.627450980392157e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 28000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.8382352941176471,
+      "grad_norm": NaN,
       "learning_rate": 3.6029411764705886e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 28500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.8529411764705882,
+      "grad_norm": NaN,
       "learning_rate": 3.5784313725490195e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 29000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.8676470588235294,
+      "grad_norm": NaN,
       "learning_rate": 3.553921568627451e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 29500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.8823529411764706,
+      "grad_norm": NaN,
       "learning_rate": 3.529411764705883e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 30000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.8970588235294118,
+      "grad_norm": NaN,
       "learning_rate": 3.5049019607843136e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 30500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.9117647058823529,
+      "grad_norm": NaN,
       "learning_rate": 3.480392156862745e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 31000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.9264705882352942,
+      "grad_norm": NaN,
       "learning_rate": 3.455882352941177e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 31500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.9411764705882353,
+      "grad_norm": NaN,
       "learning_rate": 3.431372549019608e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 32000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.9558823529411765,
+      "grad_norm": NaN,
       "learning_rate": 3.4068627450980394e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 32500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.9705882352941176,
+      "grad_norm": NaN,
       "learning_rate": 3.382352941176471e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 33000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 0.9852941176470589,
+      "grad_norm": NaN,
       "learning_rate": 3.357843137254902e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 33500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.0,
+      "grad_norm": NaN,
       "learning_rate": 3.3333333333333335e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 34000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.0147058823529411,
+      "grad_norm": NaN,
       "learning_rate": 3.308823529411765e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 34500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.0294117647058822,
+      "grad_norm": NaN,
       "learning_rate": 3.284313725490196e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 35000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.0441176470588236,
+      "grad_norm": NaN,
       "learning_rate": 3.2598039215686276e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 35500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.0588235294117647,
+      "grad_norm": NaN,
       "learning_rate": 3.235294117647059e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 36000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.0735294117647058,
+      "grad_norm": NaN,
       "learning_rate": 3.210784313725491e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 36500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.088235294117647,
+      "grad_norm": NaN,
       "learning_rate": 3.186274509803922e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 37000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.1029411764705883,
+      "grad_norm": NaN,
       "learning_rate": 3.161764705882353e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 37500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.1176470588235294,
+      "grad_norm": NaN,
       "learning_rate": 3.137254901960784e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 38000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.1323529411764706,
+      "grad_norm": NaN,
       "learning_rate": 3.112745098039216e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 38500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.1470588235294117,
+      "grad_norm": NaN,
       "learning_rate": 3.0882352941176475e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 39000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.161764705882353,
+      "grad_norm": NaN,
       "learning_rate": 3.063725490196079e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 39500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.1764705882352942,
+      "grad_norm": NaN,
       "learning_rate": 3.0392156862745097e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 40000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.1911764705882353,
+      "grad_norm": NaN,
       "learning_rate": 3.0147058823529413e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 40500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.2058823529411764,
+      "grad_norm": NaN,
       "learning_rate": 2.9901960784313725e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 41000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.2205882352941178,
+      "grad_norm": NaN,
       "learning_rate": 2.965686274509804e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 41500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.2352941176470589,
+      "grad_norm": NaN,
       "learning_rate": 2.9411764705882354e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 42000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.25,
+      "grad_norm": NaN,
       "learning_rate": 2.916666666666667e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 42500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.2647058823529411,
+      "grad_norm": NaN,
       "learning_rate": 2.8921568627450986e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 43000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.2794117647058822,
+      "grad_norm": NaN,
       "learning_rate": 2.8676470588235295e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 43500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.2941176470588236,
+      "grad_norm": NaN,
       "learning_rate": 2.8431372549019608e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 44000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.3088235294117647,
+      "grad_norm": NaN,
       "learning_rate": 2.8186274509803924e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 44500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.3235294117647058,
+      "grad_norm": NaN,
       "learning_rate": 2.7941176470588236e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 45000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.3382352941176472,
+      "grad_norm": NaN,
       "learning_rate": 2.7696078431372552e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 45500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.3529411764705883,
+      "grad_norm": NaN,
       "learning_rate": 2.7450980392156865e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 46000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.3676470588235294,
+      "grad_norm": NaN,
       "learning_rate": 2.7205882352941174e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 46500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.3823529411764706,
+      "grad_norm": NaN,
       "learning_rate": 2.696078431372549e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 47000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.3970588235294117,
+      "grad_norm": NaN,
       "learning_rate": 2.6715686274509806e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 47500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.4117647058823528,
+      "grad_norm": NaN,
       "learning_rate": 2.647058823529412e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 48000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.4264705882352942,
+      "grad_norm": NaN,
       "learning_rate": 2.6225490196078435e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 48500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.4411764705882353,
+      "grad_norm": NaN,
       "learning_rate": 2.5980392156862747e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 49000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.4558823529411764,
+      "grad_norm": NaN,
       "learning_rate": 2.5735294117647057e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 49500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.4705882352941178,
+      "grad_norm": NaN,
       "learning_rate": 2.5490196078431373e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 50000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.4852941176470589,
+      "grad_norm": NaN,
       "learning_rate": 2.5245098039215685e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 50500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.5,
+      "grad_norm": NaN,
       "learning_rate": 2.5e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 51000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.5147058823529411,
+      "grad_norm": NaN,
       "learning_rate": 2.4754901960784317e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 51500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.5294117647058822,
+      "grad_norm": NaN,
       "learning_rate": 2.4509803921568626e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 52000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.5441176470588234,
+      "grad_norm": NaN,
       "learning_rate": 2.4264705882352942e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 52500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.5588235294117647,
+      "grad_norm": NaN,
       "learning_rate": 2.401960784313726e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 53000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.5735294117647058,
+      "grad_norm": NaN,
       "learning_rate": 2.3774509803921568e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 53500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.5882352941176472,
+      "grad_norm": NaN,
       "learning_rate": 2.3529411764705884e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 54000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.6029411764705883,
+      "grad_norm": NaN,
       "learning_rate": 2.3284313725490196e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 54500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.6176470588235294,
+      "grad_norm": NaN,
       "learning_rate": 2.303921568627451e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 55000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.6323529411764706,
+      "grad_norm": NaN,
       "learning_rate": 2.2794117647058825e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 55500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.6470588235294117,
+      "grad_norm": NaN,
       "learning_rate": 2.2549019607843138e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 56000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.6617647058823528,
+      "grad_norm": NaN,
       "learning_rate": 2.2303921568627454e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 56500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.6764705882352942,
+      "grad_norm": NaN,
       "learning_rate": 2.2058823529411766e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 57000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.6911764705882353,
+      "grad_norm": NaN,
       "learning_rate": 2.181372549019608e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 57500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.7058823529411766,
+      "grad_norm": NaN,
       "learning_rate": 2.1568627450980395e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 58000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.7205882352941178,
+      "grad_norm": NaN,
       "learning_rate": 2.1323529411764707e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 58500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.7352941176470589,
+      "grad_norm": NaN,
       "learning_rate": 2.107843137254902e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 59000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.75,
+      "grad_norm": NaN,
       "learning_rate": 2.0833333333333336e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 59500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.7647058823529411,
+      "grad_norm": NaN,
       "learning_rate": 2.058823529411765e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 60000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.7794117647058822,
+      "grad_norm": NaN,
       "learning_rate": 2.034313725490196e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 60500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.7941176470588234,
+      "grad_norm": NaN,
       "learning_rate": 2.0098039215686277e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 61000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.8088235294117647,
+      "grad_norm": NaN,
       "learning_rate": 1.9852941176470586e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 61500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.8235294117647058,
+      "grad_norm": NaN,
       "learning_rate": 1.9607843137254903e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 62000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.8382352941176472,
+      "grad_norm": NaN,
       "learning_rate": 1.936274509803922e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 62500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.8529411764705883,
+      "grad_norm": NaN,
       "learning_rate": 1.9117647058823528e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 63000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.8676470588235294,
+      "grad_norm": NaN,
       "learning_rate": 1.8872549019607844e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 63500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.8823529411764706,
+      "grad_norm": NaN,
       "learning_rate": 1.862745098039216e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 64000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.8970588235294117,
+      "grad_norm": NaN,
       "learning_rate": 1.8382352941176472e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 64500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.9117647058823528,
+      "grad_norm": NaN,
       "learning_rate": 1.8137254901960785e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 65000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.9264705882352942,
+      "grad_norm": NaN,
       "learning_rate": 1.7892156862745098e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 65500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.9411764705882353,
+      "grad_norm": NaN,
       "learning_rate": 1.7647058823529414e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 66000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.9558823529411766,
+      "grad_norm": NaN,
       "learning_rate": 1.7401960784313726e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 66500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.9705882352941178,
+      "grad_norm": NaN,
       "learning_rate": 1.715686274509804e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 67000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 1.9852941176470589,
+      "grad_norm": NaN,
       "learning_rate": 1.6911764705882355e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 67500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.0,
+      "grad_norm": NaN,
       "learning_rate": 1.6666666666666667e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 68000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.014705882352941,
+      "grad_norm": NaN,
       "learning_rate": 1.642156862745098e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 68500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.0294117647058822,
+      "grad_norm": NaN,
       "learning_rate": 1.6176470588235296e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 69000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.0441176470588234,
+      "grad_norm": NaN,
       "learning_rate": 1.593137254901961e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 69500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.0588235294117645,
+      "grad_norm": NaN,
       "learning_rate": 1.568627450980392e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 70000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.073529411764706,
+      "grad_norm": NaN,
       "learning_rate": 1.5441176470588237e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 70500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.088235294117647,
+      "grad_norm": NaN,
       "learning_rate": 1.5196078431372548e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 71000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.1029411764705883,
+      "grad_norm": NaN,
       "learning_rate": 1.4950980392156863e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 71500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.1176470588235294,
+      "grad_norm": NaN,
       "learning_rate": 1.4705882352941177e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 72000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.1323529411764706,
+      "grad_norm": NaN,
       "learning_rate": 1.4460784313725493e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 72500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.1470588235294117,
+      "grad_norm": NaN,
       "learning_rate": 1.4215686274509804e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 73000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.161764705882353,
+      "grad_norm": NaN,
       "learning_rate": 1.3970588235294118e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 73500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.176470588235294,
+      "grad_norm": NaN,
       "learning_rate": 1.3725490196078432e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 74000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.1911764705882355,
+      "grad_norm": NaN,
       "learning_rate": 1.3480392156862745e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 74500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.2058823529411766,
+      "grad_norm": NaN,
       "learning_rate": 1.323529411764706e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 75000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.2205882352941178,
+      "grad_norm": NaN,
       "learning_rate": 1.2990196078431374e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 75500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.235294117647059,
+      "grad_norm": NaN,
       "learning_rate": 1.2745098039215686e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 76000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.25,
+      "grad_norm": NaN,
       "learning_rate": 1.25e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 76500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.264705882352941,
+      "grad_norm": NaN,
       "learning_rate": 1.2254901960784313e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 77000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.2794117647058822,
+      "grad_norm": NaN,
       "learning_rate": 1.200980392156863e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 77500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.2941176470588234,
+      "grad_norm": NaN,
       "learning_rate": 1.1764705882352942e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 78000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.3088235294117645,
+      "grad_norm": NaN,
       "learning_rate": 1.1519607843137254e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 78500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.323529411764706,
+      "grad_norm": NaN,
       "learning_rate": 1.1274509803921569e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 79000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.338235294117647,
+      "grad_norm": NaN,
       "learning_rate": 1.1029411764705883e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 79500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.3529411764705883,
+      "grad_norm": NaN,
       "learning_rate": 1.0784313725490197e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 80000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.3676470588235294,
+      "grad_norm": NaN,
       "learning_rate": 1.053921568627451e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 80500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.3823529411764706,
+      "grad_norm": NaN,
       "learning_rate": 1.0294117647058824e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 81000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.3970588235294117,
+      "grad_norm": NaN,
       "learning_rate": 1.0049019607843139e-05,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 81500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.411764705882353,
+      "grad_norm": NaN,
       "learning_rate": 9.803921568627451e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 82000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.426470588235294,
+      "grad_norm": NaN,
       "learning_rate": 9.558823529411764e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 82500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.4411764705882355,
+      "grad_norm": NaN,
       "learning_rate": 9.31372549019608e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 83000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.4558823529411766,
+      "grad_norm": NaN,
       "learning_rate": 9.068627450980392e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 83500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.4705882352941178,
+      "grad_norm": NaN,
       "learning_rate": 8.823529411764707e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 84000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.485294117647059,
+      "grad_norm": NaN,
       "learning_rate": 8.57843137254902e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 84500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.5,
+      "grad_norm": NaN,
       "learning_rate": 8.333333333333334e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 85000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.514705882352941,
+      "grad_norm": NaN,
       "learning_rate": 8.088235294117648e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 85500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.5294117647058822,
+      "grad_norm": NaN,
       "learning_rate": 7.84313725490196e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 86000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.5441176470588234,
+      "grad_norm": NaN,
       "learning_rate": 7.598039215686274e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 86500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.5588235294117645,
+      "grad_norm": NaN,
       "learning_rate": 7.3529411764705884e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 87000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.5735294117647056,
+      "grad_norm": NaN,
       "learning_rate": 7.107843137254902e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 87500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.588235294117647,
+      "grad_norm": NaN,
       "learning_rate": 6.862745098039216e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 88000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.6029411764705883,
+      "grad_norm": NaN,
       "learning_rate": 6.61764705882353e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 88500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.6176470588235294,
+      "grad_norm": NaN,
       "learning_rate": 6.372549019607843e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 89000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.6323529411764706,
+      "grad_norm": NaN,
       "learning_rate": 6.127450980392157e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 89500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.6470588235294117,
+      "grad_norm": NaN,
       "learning_rate": 5.882352941176471e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 90000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.661764705882353,
+      "grad_norm": NaN,
       "learning_rate": 5.637254901960784e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 90500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.6764705882352944,
+      "grad_norm": NaN,
       "learning_rate": 5.392156862745099e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 91000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.6911764705882355,
+      "grad_norm": NaN,
       "learning_rate": 5.147058823529412e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 91500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.7058823529411766,
+      "grad_norm": NaN,
       "learning_rate": 4.901960784313726e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 92000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.7205882352941178,
+      "grad_norm": NaN,
       "learning_rate": 4.65686274509804e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 92500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.735294117647059,
+      "grad_norm": NaN,
       "learning_rate": 4.411764705882353e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 93000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.75,
+      "grad_norm": NaN,
       "learning_rate": 4.166666666666667e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 93500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.764705882352941,
+      "grad_norm": NaN,
       "learning_rate": 3.92156862745098e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 94000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.7794117647058822,
+      "grad_norm": NaN,
       "learning_rate": 3.6764705882352942e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 94500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.7941176470588234,
+      "grad_norm": NaN,
       "learning_rate": 3.431372549019608e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 95000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.8088235294117645,
+      "grad_norm": NaN,
       "learning_rate": 3.1862745098039216e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 95500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.8235294117647056,
+      "grad_norm": NaN,
       "learning_rate": 2.9411764705882355e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 96000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.838235294117647,
+      "grad_norm": NaN,
       "learning_rate": 2.6960784313725493e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 96500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.8529411764705883,
+      "grad_norm": NaN,
       "learning_rate": 2.450980392156863e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 97000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.8676470588235294,
+      "grad_norm": NaN,
       "learning_rate": 2.2058823529411767e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 97500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.8823529411764706,
+      "grad_norm": NaN,
       "learning_rate": 1.96078431372549e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 98000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.8970588235294117,
+      "grad_norm": NaN,
       "learning_rate": 1.715686274509804e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 98500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.911764705882353,
+      "grad_norm": NaN,
       "learning_rate": 1.4705882352941177e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 99000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.9264705882352944,
+      "grad_norm": NaN,
       "learning_rate": 1.2254901960784314e-06,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 99500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.9411764705882355,
+      "grad_norm": NaN,
       "learning_rate": 9.80392156862745e-07,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 100000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.9558823529411766,
+      "grad_norm": NaN,
       "learning_rate": 7.352941176470589e-07,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 100500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.9705882352941178,
+      "grad_norm": NaN,
       "learning_rate": 4.901960784313725e-07,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 101000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 2.985294117647059,
+      "grad_norm": NaN,
       "learning_rate": 2.4509803921568627e-07,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 101500,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 3.0,
+      "grad_norm": NaN,
       "learning_rate": 0.0,
+      "loss": NaN,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 102000,
+      "total_memory_available (GB)": 126.62
     },
     {
       "epoch": 3.0,
+      "max_memory_allocated (GB)": 2.1,
+      "memory_allocated (GB)": 1.46,
       "step": 102000,
       "total_flos": 6.324139790696448e+19,
+      "total_memory_available (GB)": 126.62,
+      "train_loss": NaN,
+      "train_runtime": 4358.6451,
+      "train_samples_per_second": 187.213,
+      "train_steps_per_second": 23.402
     }
   ],
   "logging_steps": 500,