End of training

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 4.0,
+    "eval_accuracy": 0.9037843185753154,
+    "eval_loss": 0.4279615581035614,
+    "eval_runtime": 246.5775,
+    "eval_samples": 40430,
+    "eval_samples_per_second": 163.965,
+    "eval_steps_per_second": 5.126,
+    "train_loss": 0.16735410497358028,
+    "train_runtime": 28764.3941,
+    "train_samples": 363846,
+    "train_samples_per_second": 50.597,
+    "train_steps_per_second": 0.395
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 4.0,
+    "eval_accuracy": 0.9037843185753154,
+    "eval_loss": 0.4279615581035614,
+    "eval_runtime": 246.5775,
+    "eval_samples": 40430,
+    "eval_samples_per_second": 163.965,
+    "eval_steps_per_second": 5.126
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 4.0,
+    "train_loss": 0.16735410497358028,
+    "train_runtime": 28764.3941,
+    "train_samples": 363846,
+    "train_samples_per_second": 50.597,
+    "train_steps_per_second": 0.395
+}

trainer_state.json

@@ -0,0 +1,193 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.999120531199156,
+  "global_step": 11368,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.780524278676988e-05,
+      "loss": 0.4377,
+      "step": 500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.56060872624912e-05,
+      "loss": 0.3405,
+      "step": 1000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.340693173821253e-05,
+      "loss": 0.3114,
+      "step": 1500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.1207776213933854e-05,
+      "loss": 0.2989,
+      "step": 2000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.9008620689655174e-05,
+      "loss": 0.2909,
+      "step": 2500
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.8827108582735592,
+      "eval_loss": 0.27225127816200256,
+      "eval_runtime": 246.9326,
+      "eval_samples_per_second": 163.729,
+      "eval_steps_per_second": 5.119,
+      "step": 2842
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.68094651653765e-05,
+      "loss": 0.2555,
+      "step": 3000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.461030964109782e-05,
+      "loss": 0.2041,
+      "step": 3500
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 3.2411154116819145e-05,
+      "loss": 0.1983,
+      "step": 4000
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.0211998592540464e-05,
+      "loss": 0.1986,
+      "step": 4500
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.801284306826179e-05,
+      "loss": 0.1988,
+      "step": 5000
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 2.5813687543983113e-05,
+      "loss": 0.1941,
+      "step": 5500
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.8989116992332427,
+      "eval_loss": 0.25093942880630493,
+      "eval_runtime": 246.9753,
+      "eval_samples_per_second": 163.701,
+      "eval_steps_per_second": 5.118,
+      "step": 5685
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.3614532019704436e-05,
+      "loss": 0.1337,
+      "step": 6000
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.1419774806474316e-05,
+      "loss": 0.1001,
+      "step": 6500
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.9220619282195635e-05,
+      "loss": 0.1048,
+      "step": 7000
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.7025862068965516e-05,
+      "loss": 0.1044,
+      "step": 7500
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.4826706544686842e-05,
+      "loss": 0.1029,
+      "step": 8000
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.2627551020408163e-05,
+      "loss": 0.099,
+      "step": 8500
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.9030917635419243,
+      "eval_loss": 0.29112696647644043,
+      "eval_runtime": 246.7948,
+      "eval_samples_per_second": 163.82,
+      "eval_steps_per_second": 5.122,
+      "step": 8527
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.0428395496129487e-05,
+      "loss": 0.0457,
+      "step": 9000
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 8.22923997185081e-06,
+      "loss": 0.0428,
+      "step": 9500
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 6.030084447572133e-06,
+      "loss": 0.041,
+      "step": 10000
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 3.8309289232934554e-06,
+      "loss": 0.0372,
+      "step": 10500
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 1.6317733990147786e-06,
+      "loss": 0.0385,
+      "step": 11000
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.9037843185753154,
+      "eval_loss": 0.4279615581035614,
+      "eval_runtime": 247.0541,
+      "eval_samples_per_second": 163.648,
+      "eval_steps_per_second": 5.116,
+      "step": 11368
+    },
+    {
+      "epoch": 4.0,
+      "step": 11368,
+      "total_flos": 1.356010778555056e+18,
+      "train_loss": 0.16735410497358028,
+      "train_runtime": 28764.3941,
+      "train_samples_per_second": 50.597,
+      "train_steps_per_second": 0.395
+    }
+  ],
+  "max_steps": 11368,
+  "num_train_epochs": 4,
+  "total_flos": 1.356010778555056e+18,
+  "trial_name": null,
+  "trial_params": null
+}