End of training

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 4.0,
+    "eval_accuracy": 0.8271013754457464,
+    "eval_loss": 0.843070924282074,
+    "eval_runtime": 59.5393,
+    "eval_samples": 9815,
+    "eval_samples_per_second": 164.849,
+    "eval_steps_per_second": 5.156,
+    "train_loss": 0.28167813682183135,
+    "train_runtime": 30221.4248,
+    "train_samples": 392702,
+    "train_samples_per_second": 51.977,
+    "train_steps_per_second": 0.406
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 4.0,
+    "eval_accuracy": 0.8271013754457464,
+    "eval_loss": 0.843070924282074,
+    "eval_runtime": 59.5393,
+    "eval_samples": 9815,
+    "eval_samples_per_second": 164.849,
+    "eval_steps_per_second": 5.156
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 4.0,
+    "train_loss": 0.28167813682183135,
+    "train_runtime": 30221.4248,
+    "train_samples": 392702,
+    "train_samples_per_second": 51.977,
+    "train_steps_per_second": 0.406
+}

trainer_state.json

@@ -0,0 +1,205 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.0,
+  "global_step": 12272,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.797099087353325e-05,
+      "loss": 0.7145,
+      "step": 500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.593383311603651e-05,
+      "loss": 0.5629,
+      "step": 1000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.389667535853977e-05,
+      "loss": 0.5361,
+      "step": 1500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.185951760104303e-05,
+      "loss": 0.5212,
+      "step": 2000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.982235984354628e-05,
+      "loss": 0.5068,
+      "step": 2500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.7785202086049546e-05,
+      "loss": 0.493,
+      "step": 3000
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.8001018848700968,
+      "eval_loss": 0.5056873559951782,
+      "eval_runtime": 60.2496,
+      "eval_samples_per_second": 162.906,
+      "eval_steps_per_second": 5.095,
+      "step": 3068
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.57480443285528e-05,
+      "loss": 0.3624,
+      "step": 3500
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.3710886571056065e-05,
+      "loss": 0.3422,
+      "step": 4000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 3.167372881355932e-05,
+      "loss": 0.3507,
+      "step": 4500
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.9636571056062584e-05,
+      "loss": 0.3487,
+      "step": 5000
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.759941329856584e-05,
+      "loss": 0.3406,
+      "step": 5500
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.5562255541069103e-05,
+      "loss": 0.3426,
+      "step": 6000
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.8212939378502292,
+      "eval_loss": 0.4839480221271515,
+      "eval_runtime": 60.0624,
+      "eval_samples_per_second": 163.413,
+      "eval_steps_per_second": 5.111,
+      "step": 6136
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.3525097783572363e-05,
+      "loss": 0.2183,
+      "step": 6500
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.148794002607562e-05,
+      "loss": 0.1687,
+      "step": 7000
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.9450782268578878e-05,
+      "loss": 0.1731,
+      "step": 7500
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.7413624511082138e-05,
+      "loss": 0.1717,
+      "step": 8000
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.5376466753585397e-05,
+      "loss": 0.1664,
+      "step": 8500
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.3343383311603652e-05,
+      "loss": 0.1667,
+      "step": 9000
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.8269994905756495,
+      "eval_loss": 0.5713491439819336,
+      "eval_runtime": 60.1636,
+      "eval_samples_per_second": 163.138,
+      "eval_steps_per_second": 5.103,
+      "step": 9204
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.1306225554106911e-05,
+      "loss": 0.1069,
+      "step": 9500
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 9.26906779661017e-06,
+      "loss": 0.0606,
+      "step": 10000
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 7.235984354628423e-06,
+      "loss": 0.0587,
+      "step": 10500
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 5.198826597131682e-06,
+      "loss": 0.061,
+      "step": 11000
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 3.1616688396349417e-06,
+      "loss": 0.058,
+      "step": 11500
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 1.1245110821382008e-06,
+      "loss": 0.0537,
+      "step": 12000
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.8271013754457464,
+      "eval_loss": 0.843070924282074,
+      "eval_runtime": 60.0787,
+      "eval_samples_per_second": 163.369,
+      "eval_steps_per_second": 5.11,
+      "step": 12272
+    },
+    {
+      "epoch": 4.0,
+      "step": 12272,
+      "total_flos": 1.463867936644006e+18,
+      "train_loss": 0.28167813682183135,
+      "train_runtime": 30221.4248,
+      "train_samples_per_second": 51.977,
+      "train_steps_per_second": 0.406
+    }
+  ],
+  "max_steps": 12272,
+  "num_train_epochs": 4,
+  "total_flos": 1.463867936644006e+18,
+  "trial_name": null,
+  "trial_params": null
+}