End of training

all_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 3.93,
-    "total_flos": 5.498979435916001e+17,
-    "train_loss": 1.532381147146225,
-    "train_runtime": 193.437,
-    "train_samples_per_second": 37.221,
-    "train_steps_per_second": 0.165
+    "epoch": 0.98,
+    "total_flos": 1.3886937921591706e+17,
+    "train_loss": 1.3266394138336182,
+    "train_runtime": 44.5183,
+    "train_samples_per_second": 40.433,
+    "train_steps_per_second": 0.18
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 3.93,
-    "total_flos": 5.498979435916001e+17,
-    "train_loss": 1.532381147146225,
-    "train_runtime": 193.437,
-    "train_samples_per_second": 37.221,
-    "train_steps_per_second": 0.165
+    "epoch": 0.98,
+    "total_flos": 1.3886937921591706e+17,
+    "train_loss": 1.3266394138336182,
+    "train_runtime": 44.5183,
+    "train_samples_per_second": 40.433,
+    "train_steps_per_second": 0.18
 }

trainer_state.json

@@ -1,86 +1,38 @@
 {
-  "best_metric": 0.54,
-  "best_model_checkpoint": "alea/checkpoint-32",
-  "epoch": 3.9298245614035086,
+  "best_metric": 0.645,
+  "best_model_checkpoint": "alea/checkpoint-8",
+  "epoch": 0.9824561403508771,
   "eval_steps": 500,
-  "global_step": 32,
+  "global_step": 8,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.98,
-      "eval_accuracy": 0.295,
-      "eval_loss": 1.5737404823303223,
-      "eval_runtime": 2.351,
-      "eval_samples_per_second": 85.069,
-      "eval_steps_per_second": 2.977,
+      "eval_accuracy": 0.645,
+      "eval_loss": 1.2876272201538086,
+      "eval_runtime": 2.2928,
+      "eval_samples_per_second": 87.229,
+      "eval_steps_per_second": 3.053,
       "step": 8
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 2.310837745666504,
-      "learning_rate": 7.857142857142858e-06,
-      "loss": 1.7073,
-      "step": 10
-    },
-    {
-      "epoch": 1.96,
-      "eval_accuracy": 0.315,
-      "eval_loss": 1.5100140571594238,
-      "eval_runtime": 2.2724,
-      "eval_samples_per_second": 88.014,
-      "eval_steps_per_second": 3.081,
-      "step": 16
-    },
-    {
-      "epoch": 2.46,
-      "grad_norm": 2.365218162536621,
-      "learning_rate": 4.2857142857142855e-06,
-      "loss": 1.5052,
-      "step": 20
-    },
-    {
-      "epoch": 2.95,
-      "eval_accuracy": 0.48,
-      "eval_loss": 1.4322025775909424,
-      "eval_runtime": 2.2766,
-      "eval_samples_per_second": 87.851,
-      "eval_steps_per_second": 3.075,
-      "step": 24
-    },
-    {
-      "epoch": 3.68,
-      "grad_norm": 2.047398328781128,
-      "learning_rate": 7.142857142857143e-07,
-      "loss": 1.4107,
-      "step": 30
-    },
-    {
-      "epoch": 3.93,
-      "eval_accuracy": 0.54,
-      "eval_loss": 1.3986802101135254,
-      "eval_runtime": 2.2742,
-      "eval_samples_per_second": 87.942,
-      "eval_steps_per_second": 3.078,
-      "step": 32
-    },
-    {
-      "epoch": 3.93,
-      "step": 32,
-      "total_flos": 5.498979435916001e+17,
-      "train_loss": 1.532381147146225,
-      "train_runtime": 193.437,
-      "train_samples_per_second": 37.221,
-      "train_steps_per_second": 0.165
+      "epoch": 0.98,
+      "step": 8,
+      "total_flos": 1.3886937921591706e+17,
+      "train_loss": 1.3266394138336182,
+      "train_runtime": 44.5183,
+      "train_samples_per_second": 40.433,
+      "train_steps_per_second": 0.18
     }
   ],
   "logging_steps": 10,
-  "max_steps": 32,
+  "max_steps": 8,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 4,
+  "num_train_epochs": 1,
   "save_steps": 500,
-  "total_flos": 5.498979435916001e+17,
+  "total_flos": 1.3886937921591706e+17,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null