End of training

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 149.98,
+    "eval_loss": 0.7148692607879639,
+    "eval_runtime": 14.0171,
+    "eval_samples": 336,
+    "eval_samples_per_second": 23.971,
+    "eval_steps_per_second": 2.996,
+    "eval_wer": 0.451,
+    "train_loss": 0.9359948108413003,
+    "train_runtime": 8030.991,
+    "train_samples": 717,
+    "train_samples_per_second": 13.392,
+    "train_steps_per_second": 0.411
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 149.98,
+    "eval_loss": 0.7148692607879639,
+    "eval_runtime": 14.0171,
+    "eval_samples": 336,
+    "eval_samples_per_second": 23.971,
+    "eval_steps_per_second": 2.996,
+    "eval_wer": 0.451
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 149.98,
+    "train_loss": 0.9359948108413003,
+    "train_runtime": 8030.991,
+    "train_samples": 717,
+    "train_samples_per_second": 13.392,
+    "train_steps_per_second": 0.411
+}

trainer_state.json

@@ -0,0 +1,265 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 149.9777777777778,
+  "global_step": 3300,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 9.09,
+      "learning_rate": 4.3956000000000004e-05,
+      "loss": 9.6799,
+      "step": 200
+    },
+    {
+      "epoch": 9.09,
+      "eval_loss": 3.611884593963623,
+      "eval_runtime": 14.8755,
+      "eval_samples_per_second": 22.588,
+      "eval_steps_per_second": 2.823,
+      "eval_wer": 1.0,
+      "step": 200
+    },
+    {
+      "epoch": 18.18,
+      "learning_rate": 8.8356e-05,
+      "loss": 3.1332,
+      "step": 400
+    },
+    {
+      "epoch": 18.18,
+      "eval_loss": 2.535223960876465,
+      "eval_runtime": 14.5917,
+      "eval_samples_per_second": 23.027,
+      "eval_steps_per_second": 2.878,
+      "eval_wer": 1.005,
+      "step": 400
+    },
+    {
+      "epoch": 27.27,
+      "learning_rate": 0.000132756,
+      "loss": 1.0465,
+      "step": 600
+    },
+    {
+      "epoch": 27.27,
+      "eval_loss": 0.6169007420539856,
+      "eval_runtime": 14.7692,
+      "eval_samples_per_second": 22.75,
+      "eval_steps_per_second": 2.844,
+      "eval_wer": 0.682,
+      "step": 600
+    },
+    {
+      "epoch": 36.36,
+      "learning_rate": 0.000177156,
+      "loss": 0.3452,
+      "step": 800
+    },
+    {
+      "epoch": 36.36,
+      "eval_loss": 0.6572489142417908,
+      "eval_runtime": 14.4275,
+      "eval_samples_per_second": 23.289,
+      "eval_steps_per_second": 2.911,
+      "eval_wer": 0.607,
+      "step": 800
+    },
+    {
+      "epoch": 45.44,
+      "learning_rate": 0.000221556,
+      "loss": 0.2575,
+      "step": 1000
+    },
+    {
+      "epoch": 45.44,
+      "eval_loss": 0.6526808142662048,
+      "eval_runtime": 14.9602,
+      "eval_samples_per_second": 22.46,
+      "eval_steps_per_second": 2.807,
+      "eval_wer": 0.578,
+      "step": 1000
+    },
+    {
+      "epoch": 54.53,
+      "learning_rate": 0.0002028886956521739,
+      "loss": 0.2088,
+      "step": 1200
+    },
+    {
+      "epoch": 54.53,
+      "eval_loss": 0.6828188300132751,
+      "eval_runtime": 14.4501,
+      "eval_samples_per_second": 23.252,
+      "eval_steps_per_second": 2.907,
+      "eval_wer": 0.551,
+      "step": 1200
+    },
+    {
+      "epoch": 63.62,
+      "learning_rate": 0.00018358434782608698,
+      "loss": 0.158,
+      "step": 1400
+    },
+    {
+      "epoch": 63.62,
+      "eval_loss": 0.707419753074646,
+      "eval_runtime": 14.2371,
+      "eval_samples_per_second": 23.6,
+      "eval_steps_per_second": 2.95,
+      "eval_wer": 0.5575,
+      "step": 1400
+    },
+    {
+      "epoch": 72.71,
+      "learning_rate": 0.00016428,
+      "loss": 0.1309,
+      "step": 1600
+    },
+    {
+      "epoch": 72.71,
+      "eval_loss": 0.6523335576057434,
+      "eval_runtime": 14.2381,
+      "eval_samples_per_second": 23.599,
+      "eval_steps_per_second": 2.95,
+      "eval_wer": 0.5595,
+      "step": 1600
+    },
+    {
+      "epoch": 81.8,
+      "learning_rate": 0.00014497565217391306,
+      "loss": 0.1074,
+      "step": 1800
+    },
+    {
+      "epoch": 81.8,
+      "eval_loss": 0.7262120842933655,
+      "eval_runtime": 14.2676,
+      "eval_samples_per_second": 23.55,
+      "eval_steps_per_second": 2.944,
+      "eval_wer": 0.5415,
+      "step": 1800
+    },
+    {
+      "epoch": 90.89,
+      "learning_rate": 0.0001256713043478261,
+      "loss": 0.087,
+      "step": 2000
+    },
+    {
+      "epoch": 90.89,
+      "eval_loss": 0.7198619246482849,
+      "eval_runtime": 14.0317,
+      "eval_samples_per_second": 23.946,
+      "eval_steps_per_second": 2.993,
+      "eval_wer": 0.521,
+      "step": 2000
+    },
+    {
+      "epoch": 99.98,
+      "learning_rate": 0.00010636695652173913,
+      "loss": 0.0711,
+      "step": 2200
+    },
+    {
+      "epoch": 99.98,
+      "eval_loss": 0.7113156318664551,
+      "eval_runtime": 14.2007,
+      "eval_samples_per_second": 23.661,
+      "eval_steps_per_second": 2.958,
+      "eval_wer": 0.523,
+      "step": 2200
+    },
+    {
+      "epoch": 109.09,
+      "learning_rate": 8.706260869565218e-05,
+      "loss": 0.0601,
+      "step": 2400
+    },
+    {
+      "epoch": 109.09,
+      "eval_loss": 0.6862908601760864,
+      "eval_runtime": 13.9453,
+      "eval_samples_per_second": 24.094,
+      "eval_steps_per_second": 3.012,
+      "eval_wer": 0.496,
+      "step": 2400
+    },
+    {
+      "epoch": 118.18,
+      "learning_rate": 6.775826086956522e-05,
+      "loss": 0.0451,
+      "step": 2600
+    },
+    {
+      "epoch": 118.18,
+      "eval_loss": 0.6998496055603027,
+      "eval_runtime": 14.078,
+      "eval_samples_per_second": 23.867,
+      "eval_steps_per_second": 2.983,
+      "eval_wer": 0.483,
+      "step": 2600
+    },
+    {
+      "epoch": 127.27,
+      "learning_rate": 4.8453913043478265e-05,
+      "loss": 0.0378,
+      "step": 2800
+    },
+    {
+      "epoch": 127.27,
+      "eval_loss": 0.6970936059951782,
+      "eval_runtime": 14.3534,
+      "eval_samples_per_second": 23.409,
+      "eval_steps_per_second": 2.926,
+      "eval_wer": 0.4615,
+      "step": 2800
+    },
+    {
+      "epoch": 136.36,
+      "learning_rate": 2.9149565217391308e-05,
+      "loss": 0.0319,
+      "step": 3000
+    },
+    {
+      "epoch": 136.36,
+      "eval_loss": 0.7119385600090027,
+      "eval_runtime": 14.2911,
+      "eval_samples_per_second": 23.511,
+      "eval_steps_per_second": 2.939,
+      "eval_wer": 0.4475,
+      "step": 3000
+    },
+    {
+      "epoch": 145.44,
+      "learning_rate": 9.845217391304348e-06,
+      "loss": 0.0305,
+      "step": 3200
+    },
+    {
+      "epoch": 145.44,
+      "eval_loss": 0.7181466221809387,
+      "eval_runtime": 14.0233,
+      "eval_samples_per_second": 23.96,
+      "eval_steps_per_second": 2.995,
+      "eval_wer": 0.459,
+      "step": 3200
+    },
+    {
+      "epoch": 149.98,
+      "step": 3300,
+      "total_flos": 1.7623169369275333e+19,
+      "train_loss": 0.9359948108413003,
+      "train_runtime": 8030.991,
+      "train_samples_per_second": 13.392,
+      "train_steps_per_second": 0.411
+    }
+  ],
+  "max_steps": 3300,
+  "num_train_epochs": 150,
+  "total_flos": 1.7623169369275333e+19,
+  "trial_name": null,
+  "trial_params": null
+}