End of training

README.md

@@ -16,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model was trained from scratch on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1147.8068
+- Loss: 1150.1907
 - Wer: 1.0
 - Cer: 1.0
 

all_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 6.0,
+    "eval_cer": 1.0,
+    "eval_loss": 1150.190673828125,
+    "eval_runtime": 24.2584,
+    "eval_samples": 890,
+    "eval_samples_per_second": 36.688,
+    "eval_steps_per_second": 4.617,
+    "eval_wer": 1.0,
+    "total_flos": 5.518873820437638e+18,
+    "train_loss": 3775.4148148148147,
+    "train_runtime": 4044.046,
+    "train_samples": 17267,
+    "train_samples_per_second": 85.395,
+    "train_steps_per_second": 5.341
+}

eval--in domain_results.json

@@ -0,0 +1,10 @@
+{
+    "epoch": 6.0,
+    "eval_cer": 1.0,
+    "eval_loss": 1150.190673828125,
+    "eval_runtime": 24.2584,
+    "eval_samples": 890,
+    "eval_samples_per_second": 36.688,
+    "eval_steps_per_second": 4.617,
+    "eval_wer": 1.0
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 6.0,
+    "total_flos": 5.518873820437638e+18,
+    "train_loss": 3775.4148148148147,
+    "train_runtime": 4044.046,
+    "train_samples": 17267,
+    "train_samples_per_second": 85.395,
+    "train_steps_per_second": 5.341
+}

trainer_state.json

@@ -0,0 +1,153 @@
+{
+  "best_metric": 1.0,
+  "best_model_checkpoint": "/scratch/elec/puhe/p/palp3/sami_ASR/base_model_output/wav2vec2-base-sami-cont-pt-22k/outputs/shuff_100_no_idv/ftpseudowr25esp5/checkpoint-1080",
+  "epoch": 6.0,
+  "eval_steps": 500,
+  "global_step": 6480,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "grad_norm": 138.33615112304688,
+      "learning_rate": 9.870370370370371e-05,
+      "loss": 4573.469,
+      "step": 1080
+    },
+    {
+      "epoch": 1.0,
+      "eval_cer": 1.0,
+      "eval_loss": 1150.0875244140625,
+      "eval_runtime": 24.2162,
+      "eval_samples_per_second": 36.752,
+      "eval_steps_per_second": 4.625,
+      "eval_wer": 1.0,
+      "step": 1080
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 116.69528198242188,
+      "learning_rate": 0.00019870370370370373,
+      "loss": 3561.5266,
+      "step": 2160
+    },
+    {
+      "epoch": 2.0,
+      "eval_cer": 1.0,
+      "eval_loss": 1153.8236083984375,
+      "eval_runtime": 23.3233,
+      "eval_samples_per_second": 38.159,
+      "eval_steps_per_second": 4.802,
+      "eval_wer": 1.0,
+      "step": 2160
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 85.44125366210938,
+      "learning_rate": 0.0002987037037037037,
+      "loss": 3604.4681,
+      "step": 3240
+    },
+    {
+      "epoch": 3.0,
+      "eval_cer": 1.0,
+      "eval_loss": 1157.7269287109375,
+      "eval_runtime": 23.6203,
+      "eval_samples_per_second": 37.68,
+      "eval_steps_per_second": 4.742,
+      "eval_wer": 1.0,
+      "step": 3240
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 59.4505615234375,
+      "learning_rate": 0.00039861111111111114,
+      "loss": 3641.2648,
+      "step": 4320
+    },
+    {
+      "epoch": 4.0,
+      "eval_cer": 1.0,
+      "eval_loss": 1146.761474609375,
+      "eval_runtime": 23.5946,
+      "eval_samples_per_second": 37.72,
+      "eval_steps_per_second": 4.747,
+      "eval_wer": 1.0,
+      "step": 4320
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 150.69601440429688,
+      "learning_rate": 0.0004986111111111111,
+      "loss": 3689.1912,
+      "step": 5400
+    },
+    {
+      "epoch": 5.0,
+      "eval_cer": 1.0,
+      "eval_loss": 1146.6724853515625,
+      "eval_runtime": 23.9247,
+      "eval_samples_per_second": 37.2,
+      "eval_steps_per_second": 4.681,
+      "eval_wer": 1.0,
+      "step": 5400
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 74.24134826660156,
+      "learning_rate": 0.0004671604938271605,
+      "loss": 3582.5692,
+      "step": 6480
+    },
+    {
+      "epoch": 6.0,
+      "eval_cer": 1.0,
+      "eval_loss": 1147.8067626953125,
+      "eval_runtime": 27.5365,
+      "eval_samples_per_second": 32.321,
+      "eval_steps_per_second": 4.067,
+      "eval_wer": 1.0,
+      "step": 6480
+    },
+    {
+      "epoch": 6.0,
+      "step": 6480,
+      "total_flos": 5.518873820437638e+18,
+      "train_loss": 3775.4148148148147,
+      "train_runtime": 4044.046,
+      "train_samples_per_second": 85.395,
+      "train_steps_per_second": 5.341
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 21600,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 20,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 5,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 5
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.518873820437638e+18,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}