End of training

.ipynb_checkpoints/resume_training-checkpoint.sh

@@ -0,0 +1,33 @@
+python run_speech_recognition_ctc.py \
+	--dataset_name="mozilla-foundation/common_voice_7_0" \
+	--model_name_or_path="facebook/wav2vec2-xls-r-300m" \
+	--dataset_config_name="ar" \
+	--output_dir="./" \
+	--num_train_epochs="5" \
+	--per_device_train_batch_size="8" \
+	--per_device_eval_batch_size="8" \
+	--gradient_accumulation_steps="4" \
+	--learning_rate="7.5e-5" \
+	--warmup_steps="2000" \
+	--length_column_name="input_length" \
+	--evaluation_strategy="steps" \
+	--text_column_name="sentence" \
+	--chars_to_ignore , ? . ! \- \; \: \" “ % ‘ ” � — ’ … – \
+	--save_steps="500" \
+	--eval_steps="500" \
+	--logging_steps="100" \
+	--layerdrop="0.0" \
+	--activation_dropout="0.1" \
+	--save_total_limit="3" \
+	--freeze_feature_encoder \
+	--feat_proj_dropout="0.0" \
+	--mask_time_prob="0.75" \
+	--mask_time_length="10" \
+	--mask_feature_prob="0.25" \
+	--mask_feature_length="64" \
+	--gradient_checkpointing \
+	--use_auth_token \
+	--fp16 \
+	--group_by_length \
+	--do_train --do_eval \
+	--push_to_hub

all_results.json

@@ -1,14 +1,14 @@
 {
-    "epoch": 10.0,
-    "eval_loss": 0.4525286555290222,
-    "eval_runtime": 319.7474,
+    "epoch": 9.86,
+    "eval_loss": 0.45016008615493774,
+    "eval_runtime": 332.1791,
     "eval_samples": 10284,
-    "eval_samples_per_second": 32.163,
-    "eval_steps_per_second": 4.022,
-    "eval_wer": 0.4801227197959599,
-    "train_loss": 2.2287146272217635,
-    "train_runtime": 23537.8544,
+    "eval_samples_per_second": 30.959,
+    "eval_steps_per_second": 3.871,
+    "eval_wer": 0.4782560482007873,
+    "train_loss": 0.0,
+    "train_runtime": 3.4176,
     "train_samples": 37318,
-    "train_samples_per_second": 15.854,
-    "train_steps_per_second": 0.495
+    "train_samples_per_second": 54596.703,
+    "train_steps_per_second": 1705.873
 }

eval_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 10.0,
-    "eval_loss": 0.4525286555290222,
-    "eval_runtime": 319.7474,
+    "epoch": 9.86,
+    "eval_loss": 0.45016008615493774,
+    "eval_runtime": 332.1791,
     "eval_samples": 10284,
-    "eval_samples_per_second": 32.163,
-    "eval_steps_per_second": 4.022,
-    "eval_wer": 0.4801227197959599
+    "eval_samples_per_second": 30.959,
+    "eval_steps_per_second": 3.871,
+    "eval_wer": 0.4782560482007873
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:32a6c4cdbb67e5027c4796ff7b91aac984f0b5ba33cdc497ef815a547b4926cc
+oid sha256:023ddf68990793400757e39049470fe7511fb9501eb8e50e4a024fc333169ede
 size 1262313137

resume_training.sh

@@ -0,0 +1,33 @@
+python run_speech_recognition_ctc.py \
+	--dataset_name="mozilla-foundation/common_voice_7_0" \
+	--model_name_or_path="facebook/wav2vec2-xls-r-300m" \
+	--dataset_config_name="ar" \
+	--output_dir="./" \
+	--num_train_epochs="5" \
+	--per_device_train_batch_size="8" \
+	--per_device_eval_batch_size="8" \
+	--gradient_accumulation_steps="4" \
+	--learning_rate="7.5e-5" \
+	--warmup_steps="2000" \
+	--length_column_name="input_length" \
+	--evaluation_strategy="steps" \
+	--text_column_name="sentence" \
+	--chars_to_ignore , ? . ! \- \; \: \" “ % ‘ ” � — ’ … – \
+	--save_steps="500" \
+	--eval_steps="500" \
+	--logging_steps="100" \
+	--layerdrop="0.0" \
+	--activation_dropout="0.1" \
+	--save_total_limit="3" \
+	--freeze_feature_encoder \
+	--feat_proj_dropout="0.0" \
+	--mask_time_prob="0.75" \
+	--mask_time_length="10" \
+	--mask_feature_prob="0.25" \
+	--mask_feature_length="64" \
+	--gradient_checkpointing \
+	--use_auth_token \
+	--fp16 \
+	--group_by_length \
+	--do_train --do_eval \
+	--push_to_hub

special_tokens_map.json

@@ -1 +1 @@
-{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "[UNK]", "pad_token": "[PAD]", "additional_special_tokens": [{"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}]}
+{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "[UNK]", "pad_token": "[PAD]", "additional_special_tokens": [{"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}]}

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 10.0,
-    "train_loss": 2.2287146272217635,
-    "train_runtime": 23537.8544,
+    "epoch": 9.86,
+    "train_loss": 0.0,
+    "train_runtime": 3.4176,
     "train_samples": 37318,
-    "train_samples_per_second": 15.854,
-    "train_steps_per_second": 0.495
+    "train_samples_per_second": 54596.703,
+    "train_steps_per_second": 1705.873
 }

trainer_state.json

@@ -1,8 +1,8 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 9.99978563772776,
-  "global_step": 11660,
+  "epoch": 9.862593783494106,
+  "global_step": 11500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -905,24 +905,18 @@
       "step": 11500
     },
     {
-      "epoch": 9.95,
-      "learning_rate": 5.124223602484471e-07,
-      "loss": 1.5295,
-      "step": 11600
-    },
-    {
-      "epoch": 10.0,
-      "step": 11660,
-      "total_flos": 3.1970153172049187e+19,
-      "train_loss": 2.2287146272217635,
-      "train_runtime": 23537.8544,
-      "train_samples_per_second": 15.854,
-      "train_steps_per_second": 0.495
+      "epoch": 9.86,
+      "step": 11500,
+      "total_flos": 3.154318018894781e+19,
+      "train_loss": 0.0,
+      "train_runtime": 3.4176,
+      "train_samples_per_second": 54596.703,
+      "train_steps_per_second": 1705.873
     }
   ],
-  "max_steps": 11660,
-  "num_train_epochs": 10,
-  "total_flos": 3.1970153172049187e+19,
+  "max_steps": 5830,
+  "num_train_epochs": 5,
+  "total_flos": 3.154318018894781e+19,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8344c1dd84c70d7463dac7514ce6652d1a3d3c39cd05da2c42ca2ebd3a3a2e0b
+oid sha256:796dd8075596c17751204717a11d24106a793cf53118e126c45cc4460fd7b264
 size 2991