End of training

all_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 5.0,
+    "train_loss": 2.77278194830857,
+    "train_runtime": 10839.8418,
+    "train_samples": 42367,
+    "train_samples_per_second": 19.542,
+    "train_steps_per_second": 1.221
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 5.0,
+    "train_loss": 2.77278194830857,
+    "train_runtime": 10839.8418,
+    "train_samples": 42367,
+    "train_samples_per_second": 19.542,
+    "train_steps_per_second": 1.221
+}

train_results.txt

@@ -0,0 +1,6 @@
+epoch = 5.0
+train_loss = 2.77278194830857
+train_runtime = 10839.8418
+train_samples = 42367
+train_samples_per_second = 19.542
+train_steps_per_second = 1.221

trainer_state.json

@@ -0,0 +1,207 @@
+{
+  "best_metric": 3.903717041015625,
+  "best_model_checkpoint": "./models/gpt_test/checkpoint-10000",
+  "epoch": 5.0,
+  "global_step": 13240,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.2658610271903327e-05,
+      "loss": 3.3245,
+      "step": 1000
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 4.017566680908203,
+      "eval_runtime": 6.7548,
+      "eval_samples_per_second": 63.51,
+      "eval_steps_per_second": 3.997,
+      "step": 1000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.8298086606243708e-05,
+      "loss": 3.1222,
+      "step": 2000
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 3.984508514404297,
+      "eval_runtime": 6.7536,
+      "eval_samples_per_second": 63.522,
+      "eval_steps_per_second": 3.998,
+      "step": 2000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.5780463242698894e-05,
+      "loss": 2.9992,
+      "step": 3000
+    },
+    {
+      "epoch": 1.13,
+      "eval_loss": 3.9634785652160645,
+      "eval_runtime": 6.7624,
+      "eval_samples_per_second": 63.439,
+      "eval_steps_per_second": 3.993,
+      "step": 3000
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.3262839879154077e-05,
+      "loss": 2.8843,
+      "step": 4000
+    },
+    {
+      "epoch": 1.51,
+      "eval_loss": 3.9377200603485107,
+      "eval_runtime": 6.7434,
+      "eval_samples_per_second": 63.618,
+      "eval_steps_per_second": 4.004,
+      "step": 4000
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 2.0745216515609266e-05,
+      "loss": 2.882,
+      "step": 5000
+    },
+    {
+      "epoch": 1.89,
+      "eval_loss": 3.926788091659546,
+      "eval_runtime": 6.7619,
+      "eval_samples_per_second": 63.443,
+      "eval_steps_per_second": 3.993,
+      "step": 5000
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.822759315206445e-05,
+      "loss": 2.7411,
+      "step": 6000
+    },
+    {
+      "epoch": 2.27,
+      "eval_loss": 3.9208364486694336,
+      "eval_runtime": 6.7546,
+      "eval_samples_per_second": 63.512,
+      "eval_steps_per_second": 3.997,
+      "step": 6000
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.570996978851964e-05,
+      "loss": 2.7204,
+      "step": 7000
+    },
+    {
+      "epoch": 2.64,
+      "eval_loss": 3.916015148162842,
+      "eval_runtime": 6.7585,
+      "eval_samples_per_second": 63.476,
+      "eval_steps_per_second": 3.995,
+      "step": 7000
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.3192346424974824e-05,
+      "loss": 2.7106,
+      "step": 8000
+    },
+    {
+      "epoch": 3.02,
+      "eval_loss": 3.9170773029327393,
+      "eval_runtime": 6.749,
+      "eval_samples_per_second": 63.565,
+      "eval_steps_per_second": 4.001,
+      "step": 8000
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 1.067472306143001e-05,
+      "loss": 2.5857,
+      "step": 9000
+    },
+    {
+      "epoch": 3.4,
+      "eval_loss": 3.9161624908447266,
+      "eval_runtime": 6.7583,
+      "eval_samples_per_second": 63.477,
+      "eval_steps_per_second": 3.995,
+      "step": 9000
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 8.157099697885196e-06,
+      "loss": 2.5863,
+      "step": 10000
+    },
+    {
+      "epoch": 3.78,
+      "eval_loss": 3.903717041015625,
+      "eval_runtime": 6.7566,
+      "eval_samples_per_second": 63.493,
+      "eval_steps_per_second": 3.996,
+      "step": 10000
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 5.6394763343403835e-06,
+      "loss": 2.5674,
+      "step": 11000
+    },
+    {
+      "epoch": 4.15,
+      "eval_loss": 3.9134981632232666,
+      "eval_runtime": 6.7557,
+      "eval_samples_per_second": 63.502,
+      "eval_steps_per_second": 3.997,
+      "step": 11000
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 3.121852970795569e-06,
+      "loss": 2.4901,
+      "step": 12000
+    },
+    {
+      "epoch": 4.53,
+      "eval_loss": 3.9125208854675293,
+      "eval_runtime": 6.7457,
+      "eval_samples_per_second": 63.596,
+      "eval_steps_per_second": 4.003,
+      "step": 12000
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 6.042296072507553e-07,
+      "loss": 2.505,
+      "step": 13000
+    },
+    {
+      "epoch": 4.91,
+      "eval_loss": 3.9078235626220703,
+      "eval_runtime": 6.747,
+      "eval_samples_per_second": 63.584,
+      "eval_steps_per_second": 4.002,
+      "step": 13000
+    },
+    {
+      "epoch": 5.0,
+      "step": 13240,
+      "total_flos": 5.535079759872e+16,
+      "train_loss": 2.77278194830857,
+      "train_runtime": 10839.8418,
+      "train_samples_per_second": 19.542,
+      "train_steps_per_second": 1.221
+    }
+  ],
+  "max_steps": 13240,
+  "num_train_epochs": 5,
+  "total_flos": 5.535079759872e+16,
+  "trial_name": null,
+  "trial_params": null
+}