End of training

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 4.0,
+    "eval_accuracy": 0.9116497650259708,
+    "eval_loss": 0.2742404043674469,
+    "eval_runtime": 227.3198,
+    "eval_samples": 40430,
+    "eval_samples_per_second": 177.855,
+    "eval_steps_per_second": 5.56,
+    "train_loss": 0.19540013602550058,
+    "train_runtime": 25859.9259,
+    "train_samples": 363846,
+    "train_samples_per_second": 56.28,
+    "train_steps_per_second": 0.44
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 4.0,
+    "eval_accuracy": 0.9116497650259708,
+    "eval_loss": 0.2742404043674469,
+    "eval_runtime": 227.3198,
+    "eval_samples": 40430,
+    "eval_samples_per_second": 177.855,
+    "eval_steps_per_second": 5.56
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 4.0,
+    "train_loss": 0.19540013602550058,
+    "train_runtime": 25859.9259,
+    "train_samples": 363846,
+    "train_samples_per_second": 56.28,
+    "train_steps_per_second": 0.44
+}

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": 4.7822836030964113e-05,
+      "loss": 0.452,
+      "step": 500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.5628078817733994e-05,
+      "loss": 0.335,
+      "step": 1000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.342892329345531e-05,
+      "loss": 0.3035,
+      "step": 1500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.1238564391273755e-05,
+      "loss": 0.3312,
+      "step": 2000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.9039408866995074e-05,
+      "loss": 0.2866,
+      "step": 2500
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.8891169923324265,
+      "eval_loss": 0.2588575780391693,
+      "eval_runtime": 228.3833,
+      "eval_samples_per_second": 177.027,
+      "eval_steps_per_second": 5.535,
+      "step": 2842
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.684025334271639e-05,
+      "loss": 0.2468,
+      "step": 3000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.4641097818437726e-05,
+      "loss": 0.2033,
+      "step": 3500
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 3.2441942294159045e-05,
+      "loss": 0.249,
+      "step": 4000
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.0242786769880365e-05,
+      "loss": 0.2109,
+      "step": 4500
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.8043631245601694e-05,
+      "loss": 0.2063,
+      "step": 5000
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 2.5844475721323013e-05,
+      "loss": 0.2022,
+      "step": 5500
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.8969577046747464,
+      "eval_loss": 0.2509480118751526,
+      "eval_runtime": 228.2046,
+      "eval_samples_per_second": 177.166,
+      "eval_steps_per_second": 5.539,
+      "step": 5685
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.3645320197044336e-05,
+      "loss": 0.1695,
+      "step": 6000
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.144616467276566e-05,
+      "loss": 0.1429,
+      "step": 6500
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.924700914848698e-05,
+      "loss": 0.1449,
+      "step": 7000
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.7047853624208304e-05,
+      "loss": 0.143,
+      "step": 7500
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.4848698099929629e-05,
+      "loss": 0.1337,
+      "step": 8000
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.264954257565095e-05,
+      "loss": 0.1383,
+      "step": 8500
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.9082611921840218,
+      "eval_loss": 0.2720527648925781,
+      "eval_runtime": 228.115,
+      "eval_samples_per_second": 177.235,
+      "eval_steps_per_second": 5.541,
+      "step": 8527
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.0450387051372274e-05,
+      "loss": 0.0984,
+      "step": 9000
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 8.251231527093597e-06,
+      "loss": 0.094,
+      "step": 9500
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 6.052076002814919e-06,
+      "loss": 0.0955,
+      "step": 10000
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 3.852920478536242e-06,
+      "loss": 0.095,
+      "step": 10500
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 1.6537649542575654e-06,
+      "loss": 0.0938,
+      "step": 11000
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.9116497650259708,
+      "eval_loss": 0.2742404043674469,
+      "eval_runtime": 228.0285,
+      "eval_samples_per_second": 177.302,
+      "eval_steps_per_second": 5.543,
+      "step": 11368
+    },
+    {
+      "epoch": 4.0,
+      "step": 11368,
+      "total_flos": 1.3560290965631263e+18,
+      "train_loss": 0.19540013602550058,
+      "train_runtime": 25859.9259,
+      "train_samples_per_second": 56.28,
+      "train_steps_per_second": 0.44
+    }
+  ],
+  "max_steps": 11368,
+  "num_train_epochs": 4,
+  "total_flos": 1.3560290965631263e+18,
+  "trial_name": null,
+  "trial_params": null
+}