End of training

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 4.0,
+    "eval_accuracy": 0.8202750891492613,
+    "eval_loss": 0.8249099850654602,
+    "eval_runtime": 21.0971,
+    "eval_samples": 9815,
+    "eval_samples_per_second": 465.231,
+    "eval_steps_per_second": 14.552,
+    "train_loss": 0.2962776091450041,
+    "train_runtime": 10975.1929,
+    "train_samples": 392702,
+    "train_samples_per_second": 143.123,
+    "train_steps_per_second": 1.118
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 4.0,
+    "eval_accuracy": 0.8202750891492613,
+    "eval_loss": 0.8249099850654602,
+    "eval_runtime": 21.0971,
+    "eval_samples": 9815,
+    "eval_samples_per_second": 465.231,
+    "eval_steps_per_second": 14.552
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 4.0,
+    "train_loss": 0.2962776091450041,
+    "train_runtime": 10975.1929,
+    "train_samples": 392702,
+    "train_samples_per_second": 143.123,
+    "train_steps_per_second": 1.118
+}

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": 4.796691655801825e-05,
+      "loss": 0.6778,
+      "step": 500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.5941981747066495e-05,
+      "loss": 0.5718,
+      "step": 1000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.390482398956976e-05,
+      "loss": 0.5419,
+      "step": 1500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.1867666232073014e-05,
+      "loss": 0.5238,
+      "step": 2000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.983050847457627e-05,
+      "loss": 0.512,
+      "step": 2500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.779335071707953e-05,
+      "loss": 0.4968,
+      "step": 3000
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.8080489047376465,
+      "eval_loss": 0.4887898862361908,
+      "eval_runtime": 21.3558,
+      "eval_samples_per_second": 459.594,
+      "eval_steps_per_second": 14.375,
+      "step": 3068
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.576026727509779e-05,
+      "loss": 0.3678,
+      "step": 3500
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.3723109517601045e-05,
+      "loss": 0.353,
+      "step": 4000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 3.16859517601043e-05,
+      "loss": 0.3553,
+      "step": 4500
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.9648794002607567e-05,
+      "loss": 0.3544,
+      "step": 5000
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.7611636245110823e-05,
+      "loss": 0.3569,
+      "step": 5500
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.5574478487614083e-05,
+      "loss": 0.3496,
+      "step": 6000
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.8190524707080998,
+      "eval_loss": 0.49091723561286926,
+      "eval_runtime": 21.2096,
+      "eval_samples_per_second": 462.762,
+      "eval_steps_per_second": 14.475,
+      "step": 6136
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.353732073011734e-05,
+      "loss": 0.2338,
+      "step": 6500
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.15001629726206e-05,
+      "loss": 0.1957,
+      "step": 7000
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.946300521512386e-05,
+      "loss": 0.1952,
+      "step": 7500
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.742584745762712e-05,
+      "loss": 0.1935,
+      "step": 8000
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.538868970013038e-05,
+      "loss": 0.1953,
+      "step": 8500
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.3355606258148632e-05,
+      "loss": 0.1924,
+      "step": 9000
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.8198675496688742,
+      "eval_loss": 0.6029134392738342,
+      "eval_runtime": 21.1893,
+      "eval_samples_per_second": 463.206,
+      "eval_steps_per_second": 14.488,
+      "step": 9204
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.1318448500651891e-05,
+      "loss": 0.1305,
+      "step": 9500
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 9.28129074315515e-06,
+      "loss": 0.0882,
+      "step": 10000
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 7.24413298565841e-06,
+      "loss": 0.0921,
+      "step": 10500
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 5.211049543676663e-06,
+      "loss": 0.0855,
+      "step": 11000
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 3.1738917861799224e-06,
+      "loss": 0.0828,
+      "step": 11500
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 1.1367340286831812e-06,
+      "loss": 0.0811,
+      "step": 12000
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.8202750891492613,
+      "eval_loss": 0.8249099850654602,
+      "eval_runtime": 21.2346,
+      "eval_samples_per_second": 462.217,
+      "eval_steps_per_second": 14.458,
+      "step": 12272
+    },
+    {
+      "epoch": 4.0,
+      "step": 12272,
+      "total_flos": 4.104507322049495e+17,
+      "train_loss": 0.2962776091450041,
+      "train_runtime": 10975.1929,
+      "train_samples_per_second": 143.123,
+      "train_steps_per_second": 1.118
+    }
+  ],
+  "max_steps": 12272,
+  "num_train_epochs": 4,
+  "total_flos": 4.104507322049495e+17,
+  "trial_name": null,
+  "trial_params": null
+}