End of training

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 200.0,
+    "eval_loss": 0.8880954384803772,
+    "eval_runtime": 13.3973,
+    "eval_samples": 360,
+    "eval_samples_per_second": 26.871,
+    "eval_steps_per_second": 0.896,
+    "eval_wer": 0.4174833398667189,
+    "train_loss": 1.469863431012189,
+    "train_runtime": 10503.1195,
+    "train_samples": 856,
+    "train_samples_per_second": 16.3,
+    "train_steps_per_second": 0.514
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 200.0,
+    "eval_loss": 0.8880954384803772,
+    "eval_runtime": 13.3973,
+    "eval_samples": 360,
+    "eval_samples_per_second": 26.871,
+    "eval_steps_per_second": 0.896,
+    "eval_wer": 0.4174833398667189
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 200.0,
+    "train_loss": 1.469863431012189,
+    "train_runtime": 10503.1195,
+    "train_samples": 856,
+    "train_samples_per_second": 16.3,
+    "train_steps_per_second": 0.514
+}

trainer_state.json

@@ -0,0 +1,175 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 200.0,
+  "global_step": 5400,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 18.52,
+      "learning_rate": 2.7639e-05,
+      "loss": 10.695,
+      "step": 500
+    },
+    {
+      "epoch": 18.52,
+      "eval_loss": 3.5681302547454834,
+      "eval_runtime": 13.5335,
+      "eval_samples_per_second": 26.601,
+      "eval_steps_per_second": 0.887,
+      "eval_wer": 1.0,
+      "step": 500
+    },
+    {
+      "epoch": 37.04,
+      "learning_rate": 5.5389e-05,
+      "loss": 3.2718,
+      "step": 1000
+    },
+    {
+      "epoch": 37.04,
+      "eval_loss": 2.308128833770752,
+      "eval_runtime": 13.213,
+      "eval_samples_per_second": 27.246,
+      "eval_steps_per_second": 0.908,
+      "eval_wer": 0.964327714621717,
+      "step": 1000
+    },
+    {
+      "epoch": 55.56,
+      "learning_rate": 8.3139e-05,
+      "loss": 0.8727,
+      "step": 1500
+    },
+    {
+      "epoch": 55.56,
+      "eval_loss": 0.7227439284324646,
+      "eval_runtime": 13.6912,
+      "eval_samples_per_second": 26.294,
+      "eval_steps_per_second": 0.876,
+      "eval_wer": 0.5147001176009408,
+      "step": 1500
+    },
+    {
+      "epoch": 74.07,
+      "learning_rate": 0.000110889,
+      "loss": 0.3349,
+      "step": 2000
+    },
+    {
+      "epoch": 74.07,
+      "eval_loss": 0.7497692704200745,
+      "eval_runtime": 13.3287,
+      "eval_samples_per_second": 27.009,
+      "eval_steps_per_second": 0.9,
+      "eval_wer": 0.4958839670717366,
+      "step": 2000
+    },
+    {
+      "epoch": 92.59,
+      "learning_rate": 9.474176470588235e-05,
+      "loss": 0.2134,
+      "step": 2500
+    },
+    {
+      "epoch": 92.59,
+      "eval_loss": 0.7779262065887451,
+      "eval_runtime": 13.3535,
+      "eval_samples_per_second": 26.959,
+      "eval_steps_per_second": 0.899,
+      "eval_wer": 0.4719717757742062,
+      "step": 2500
+    },
+    {
+      "epoch": 111.11,
+      "learning_rate": 7.841823529411764e-05,
+      "loss": 0.1445,
+      "step": 3000
+    },
+    {
+      "epoch": 111.11,
+      "eval_loss": 0.812047004699707,
+      "eval_runtime": 13.2296,
+      "eval_samples_per_second": 27.212,
+      "eval_steps_per_second": 0.907,
+      "eval_wer": 0.4594276754214034,
+      "step": 3000
+    },
+    {
+      "epoch": 129.63,
+      "learning_rate": 6.209470588235294e-05,
+      "loss": 0.1057,
+      "step": 3500
+    },
+    {
+      "epoch": 129.63,
+      "eval_loss": 0.8224607706069946,
+      "eval_runtime": 13.3049,
+      "eval_samples_per_second": 27.058,
+      "eval_steps_per_second": 0.902,
+      "eval_wer": 0.4609956879655037,
+      "step": 3500
+    },
+    {
+      "epoch": 148.15,
+      "learning_rate": 4.577117647058824e-05,
+      "loss": 0.0826,
+      "step": 4000
+    },
+    {
+      "epoch": 148.15,
+      "eval_loss": 0.8306752443313599,
+      "eval_runtime": 13.3606,
+      "eval_samples_per_second": 26.945,
+      "eval_steps_per_second": 0.898,
+      "eval_wer": 0.4351234809878479,
+      "step": 4000
+    },
+    {
+      "epoch": 166.67,
+      "learning_rate": 2.9447647058823532e-05,
+      "loss": 0.0639,
+      "step": 4500
+    },
+    {
+      "epoch": 166.67,
+      "eval_loss": 0.8967453241348267,
+      "eval_runtime": 13.2264,
+      "eval_samples_per_second": 27.218,
+      "eval_steps_per_second": 0.907,
+      "eval_wer": 0.4315954527636221,
+      "step": 4500
+    },
+    {
+      "epoch": 185.19,
+      "learning_rate": 1.3124117647058824e-05,
+      "loss": 0.0528,
+      "step": 5000
+    },
+    {
+      "epoch": 185.19,
+      "eval_loss": 0.8874959349632263,
+      "eval_runtime": 13.2295,
+      "eval_samples_per_second": 27.212,
+      "eval_steps_per_second": 0.907,
+      "eval_wer": 0.42375539004312035,
+      "step": 5000
+    },
+    {
+      "epoch": 200.0,
+      "step": 5400,
+      "total_flos": 2.7163202874536632e+19,
+      "train_loss": 1.469863431012189,
+      "train_runtime": 10503.1195,
+      "train_samples_per_second": 16.3,
+      "train_steps_per_second": 0.514
+    }
+  ],
+  "max_steps": 5400,
+  "num_train_epochs": 200,
+  "total_flos": 2.7163202874536632e+19,
+  "trial_name": null,
+  "trial_params": null
+}