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: 1151.3530
+- Loss: 1150.0363
 - Wer: 1.0
 - Cer: 1.0
 

all_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 6.0,
+    "eval_cer": 1.0,
+    "eval_loss": 1150.0362548828125,
+    "eval_runtime": 23.9604,
+    "eval_samples": 890,
+    "eval_samples_per_second": 37.145,
+    "eval_steps_per_second": 4.674,
+    "eval_wer": 1.0,
+    "total_flos": 5.519100286987747e+18,
+    "train_loss": 3828.2266203703703,
+    "train_runtime": 3861.8113,
+    "train_samples": 17267,
+    "train_samples_per_second": 89.424,
+    "train_steps_per_second": 5.593
+}

eval--in domain_results.json

@@ -0,0 +1,10 @@
+{
+    "epoch": 6.0,
+    "eval_cer": 1.0,
+    "eval_loss": 1150.0362548828125,
+    "eval_runtime": 23.9604,
+    "eval_samples": 890,
+    "eval_samples_per_second": 37.145,
+    "eval_steps_per_second": 4.674,
+    "eval_wer": 1.0
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 6.0,
+    "total_flos": 5.519100286987747e+18,
+    "train_loss": 3828.2266203703703,
+    "train_runtime": 3861.8113,
+    "train_samples": 17267,
+    "train_samples_per_second": 89.424,
+    "train_steps_per_second": 5.593
+}

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_142/ftpseudowr20esp5/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": 110.69710540771484,
+      "learning_rate": 0.00012337962962962964,
+      "loss": 4511.6611,
+      "step": 1080
+    },
+    {
+      "epoch": 1.0,
+      "eval_cer": 1.0,
+      "eval_loss": 1149.935302734375,
+      "eval_runtime": 23.6152,
+      "eval_samples_per_second": 37.688,
+      "eval_steps_per_second": 4.743,
+      "eval_wer": 1.0,
+      "step": 1080
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 134.31155395507812,
+      "learning_rate": 0.00024837962962962964,
+      "loss": 3591.4514,
+      "step": 2160
+    },
+    {
+      "epoch": 2.0,
+      "eval_cer": 1.0,
+      "eval_loss": 1155.4658203125,
+      "eval_runtime": 23.3518,
+      "eval_samples_per_second": 38.113,
+      "eval_steps_per_second": 4.796,
+      "eval_wer": 1.0,
+      "step": 2160
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 156.53646850585938,
+      "learning_rate": 0.00037337962962962965,
+      "loss": 3649.9262,
+      "step": 3240
+    },
+    {
+      "epoch": 3.0,
+      "eval_cer": 1.0,
+      "eval_loss": 1155.608154296875,
+      "eval_runtime": 23.3593,
+      "eval_samples_per_second": 38.1,
+      "eval_steps_per_second": 4.795,
+      "eval_wer": 1.0,
+      "step": 3240
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 100.56692504882812,
+      "learning_rate": 0.0004982638888888889,
+      "loss": 3789.6616,
+      "step": 4320
+    },
+    {
+      "epoch": 4.0,
+      "eval_cer": 1.0,
+      "eval_loss": 1146.7283935546875,
+      "eval_runtime": 23.3466,
+      "eval_samples_per_second": 38.121,
+      "eval_steps_per_second": 4.797,
+      "eval_wer": 1.0,
+      "step": 4320
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 103.7132568359375,
+      "learning_rate": 0.0004691840277777778,
+      "loss": 3752.8502,
+      "step": 5400
+    },
+    {
+      "epoch": 5.0,
+      "eval_cer": 1.0,
+      "eval_loss": 1158.9339599609375,
+      "eval_runtime": 26.3666,
+      "eval_samples_per_second": 33.755,
+      "eval_steps_per_second": 4.248,
+      "eval_wer": 1.0,
+      "step": 5400
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 97.61384582519531,
+      "learning_rate": 0.0004379340277777778,
+      "loss": 3673.8093,
+      "step": 6480
+    },
+    {
+      "epoch": 6.0,
+      "eval_cer": 1.0,
+      "eval_loss": 1151.35302734375,
+      "eval_runtime": 25.4193,
+      "eval_samples_per_second": 35.013,
+      "eval_steps_per_second": 4.406,
+      "eval_wer": 1.0,
+      "step": 6480
+    },
+    {
+      "epoch": 6.0,
+      "step": 6480,
+      "total_flos": 5.519100286987747e+18,
+      "train_loss": 3828.2266203703703,
+      "train_runtime": 3861.8113,
+      "train_samples_per_second": 89.424,
+      "train_steps_per_second": 5.593
+    }
+  ],
+  "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": 5.519100286987747e+18,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}