End of training

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 15.0,
+    "eval_loss": 0.5528259873390198,
+    "eval_runtime": 49.5497,
+    "eval_samples": 2027,
+    "eval_samples_per_second": 40.908,
+    "eval_steps_per_second": 5.126,
+    "eval_wer": 0.38111572486008305,
+    "train_loss": 1.1127878371407003,
+    "train_runtime": 3384.281,
+    "train_samples": 4350,
+    "train_samples_per_second": 19.28,
+    "train_steps_per_second": 0.603
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 15.0,
+    "eval_loss": 0.5528259873390198,
+    "eval_runtime": 49.5497,
+    "eval_samples": 2027,
+    "eval_samples_per_second": 40.908,
+    "eval_steps_per_second": 5.126,
+    "eval_wer": 0.38111572486008305
+}

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dcd7bf2f05da5d16f6f8ada9e0fa05f844ed883a19f72bb2a39311f1fa21c174
+oid sha256:e4359cf5e096bfe5f441965025547f3101bdd623f9afa5d21c45e02babd0682b
 size 1262063089

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 15.0,
+    "train_loss": 1.1127878371407003,
+    "train_runtime": 3384.281,
+    "train_samples": 4350,
+    "train_samples_per_second": 19.28,
+    "train_steps_per_second": 0.603
+}

trainer_state.json

@@ -0,0 +1,229 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 15.0,
+  "global_step": 2040,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.74,
+      "eval_loss": 3.44441294670105,
+      "eval_runtime": 49.9028,
+      "eval_samples_per_second": 40.619,
+      "eval_steps_per_second": 5.09,
+      "eval_wer": 1.0,
+      "step": 100
+    },
+    {
+      "epoch": 1.47,
+      "eval_loss": 2.942133903503418,
+      "eval_runtime": 49.8952,
+      "eval_samples_per_second": 40.625,
+      "eval_steps_per_second": 5.091,
+      "eval_wer": 1.0,
+      "step": 200
+    },
+    {
+      "epoch": 2.21,
+      "eval_loss": 2.280221939086914,
+      "eval_runtime": 49.5318,
+      "eval_samples_per_second": 40.923,
+      "eval_steps_per_second": 5.128,
+      "eval_wer": 1.013720888246976,
+      "step": 300
+    },
+    {
+      "epoch": 2.94,
+      "eval_loss": 0.9682899117469788,
+      "eval_runtime": 48.9178,
+      "eval_samples_per_second": 41.437,
+      "eval_steps_per_second": 5.192,
+      "eval_wer": 0.7610579526990432,
+      "step": 400
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 0.0002982,
+      "loss": 3.7264,
+      "step": 500
+    },
+    {
+      "epoch": 3.68,
+      "eval_loss": 0.7941266298294067,
+      "eval_runtime": 49.6243,
+      "eval_samples_per_second": 40.847,
+      "eval_steps_per_second": 5.118,
+      "eval_wer": 0.659415056869471,
+      "step": 500
+    },
+    {
+      "epoch": 4.41,
+      "eval_loss": 0.6694735884666443,
+      "eval_runtime": 48.7117,
+      "eval_samples_per_second": 41.612,
+      "eval_steps_per_second": 5.214,
+      "eval_wer": 0.5751038093518686,
+      "step": 600
+    },
+    {
+      "epoch": 5.15,
+      "eval_loss": 0.6507321000099182,
+      "eval_runtime": 49.0122,
+      "eval_samples_per_second": 41.357,
+      "eval_steps_per_second": 5.182,
+      "eval_wer": 0.5314136125654451,
+      "step": 700
+    },
+    {
+      "epoch": 5.88,
+      "eval_loss": 0.5731020569801331,
+      "eval_runtime": 48.7959,
+      "eval_samples_per_second": 41.54,
+      "eval_steps_per_second": 5.205,
+      "eval_wer": 0.4926882108683878,
+      "step": 800
+    },
+    {
+      "epoch": 6.62,
+      "eval_loss": 0.5722740292549133,
+      "eval_runtime": 49.263,
+      "eval_samples_per_second": 41.147,
+      "eval_steps_per_second": 5.156,
+      "eval_wer": 0.45802491424444847,
+      "step": 900
+    },
+    {
+      "epoch": 7.35,
+      "learning_rate": 0.00020318181818181815,
+      "loss": 0.4592,
+      "step": 1000
+    },
+    {
+      "epoch": 7.35,
+      "eval_loss": 0.5912505388259888,
+      "eval_runtime": 49.0337,
+      "eval_samples_per_second": 41.339,
+      "eval_steps_per_second": 5.18,
+      "eval_wer": 0.44791478606246615,
+      "step": 1000
+    },
+    {
+      "epoch": 8.09,
+      "eval_loss": 0.5561510324478149,
+      "eval_runtime": 49.6104,
+      "eval_samples_per_second": 40.858,
+      "eval_steps_per_second": 5.12,
+      "eval_wer": 0.44231810796172594,
+      "step": 1100
+    },
+    {
+      "epoch": 8.82,
+      "eval_loss": 0.5565929412841797,
+      "eval_runtime": 48.96,
+      "eval_samples_per_second": 41.401,
+      "eval_steps_per_second": 5.188,
+      "eval_wer": 0.42922910272612386,
+      "step": 1200
+    },
+    {
+      "epoch": 9.56,
+      "eval_loss": 0.5492243766784668,
+      "eval_runtime": 49.7971,
+      "eval_samples_per_second": 40.705,
+      "eval_steps_per_second": 5.101,
+      "eval_wer": 0.430312330745622,
+      "step": 1300
+    },
+    {
+      "epoch": 10.29,
+      "eval_loss": 0.5664941668510437,
+      "eval_runtime": 49.6868,
+      "eval_samples_per_second": 40.796,
+      "eval_steps_per_second": 5.112,
+      "eval_wer": 0.43311066979599205,
+      "step": 1400
+    },
+    {
+      "epoch": 11.03,
+      "learning_rate": 0.00010577922077922077,
+      "loss": 0.2121,
+      "step": 1500
+    },
+    {
+      "epoch": 11.03,
+      "eval_loss": 0.5609949231147766,
+      "eval_runtime": 49.3464,
+      "eval_samples_per_second": 41.077,
+      "eval_steps_per_second": 5.147,
+      "eval_wer": 0.4083769633507853,
+      "step": 1500
+    },
+    {
+      "epoch": 11.76,
+      "eval_loss": 0.5702986717224121,
+      "eval_runtime": 49.4945,
+      "eval_samples_per_second": 40.954,
+      "eval_steps_per_second": 5.132,
+      "eval_wer": 0.4014262502256725,
+      "step": 1600
+    },
+    {
+      "epoch": 12.5,
+      "eval_loss": 0.5668947696685791,
+      "eval_runtime": 49.1446,
+      "eval_samples_per_second": 41.246,
+      "eval_steps_per_second": 5.168,
+      "eval_wer": 0.3897815490160679,
+      "step": 1700
+    },
+    {
+      "epoch": 13.24,
+      "eval_loss": 0.5585775971412659,
+      "eval_runtime": 49.1158,
+      "eval_samples_per_second": 41.27,
+      "eval_steps_per_second": 5.171,
+      "eval_wer": 0.3961906481314317,
+      "step": 1800
+    },
+    {
+      "epoch": 13.97,
+      "eval_loss": 0.5655893683433533,
+      "eval_runtime": 49.1541,
+      "eval_samples_per_second": 41.238,
+      "eval_steps_per_second": 5.167,
+      "eval_wer": 0.38969128001444303,
+      "step": 1900
+    },
+    {
+      "epoch": 14.71,
+      "learning_rate": 8.376623376623376e-06,
+      "loss": 0.1326,
+      "step": 2000
+    },
+    {
+      "epoch": 14.71,
+      "eval_loss": 0.5565294027328491,
+      "eval_runtime": 48.8212,
+      "eval_samples_per_second": 41.519,
+      "eval_steps_per_second": 5.203,
+      "eval_wer": 0.38129626286333274,
+      "step": 2000
+    },
+    {
+      "epoch": 15.0,
+      "step": 2040,
+      "total_flos": 6.61647473178647e+18,
+      "train_loss": 1.1127878371407003,
+      "train_runtime": 3384.281,
+      "train_samples_per_second": 19.28,
+      "train_steps_per_second": 0.603
+    }
+  ],
+  "max_steps": 2040,
+  "num_train_epochs": 15,
+  "total_flos": 6.61647473178647e+18,
+  "trial_name": null,
+  "trial_params": null
+}