Training in progress, step 500

.ipynb_checkpoints/run-checkpoint.sh

@@ -13,7 +13,7 @@ python xls-r-uzbek-cv8/run_speech_recognition_ctc.py \
 	--length_column_name="input_length" \
 	--evaluation_strategy="steps" \
 	--text_column_name="sentence" \
-	--eval_metrics="wer cer" \
+	--eval_metrics="cer" \
 	--save_steps="500" \
 	--eval_steps="500" \
 	--logging_steps="100" \

.ipynb_checkpoints/run_speech_recognition_ctc-checkpoint.py

@@ -448,7 +448,11 @@ def main():
             batch["target_text"] = re.sub(
                 chars_to_ignore_regex, 
                 "", 
-                re.sub("([og])['`´]", "\g<1>‘", unidecode.unidecode(batch[text_column_name]).lower())
+                re.sub("['`´]", "’",   # elsewhere probably meant as glottal stop
+                       re.sub("([og])['`´]", "\g<1>‘",  # after o/g indicate modified char
+                              unidecode.unidecode(batch[text_column_name]).lower()
+                             )
+                      )
             ) + " "
         else:
             batch["target_text"] = batch[text_column_name].lower() + " "

added_tokens.json

@@ -1 +1 @@
-{"<s>": 30, "</s>": 31}
+{"<s>": 31, "</s>": 32}

config.json

@@ -76,7 +76,7 @@
   "num_hidden_layers": 24,
   "num_negatives": 100,
   "output_hidden_size": 1024,
-  "pad_token_id": 29,
+  "pad_token_id": 30,
   "proj_codevector_dim": 768,
   "tdnn_dilation": [
     1,
@@ -102,7 +102,7 @@
   "torch_dtype": "float32",
   "transformers_version": "4.17.0.dev0",
   "use_weighted_layer_sum": false,
-  "vocab_size": 32,
+  "vocab_size": 33,
   "xvector_output_dim": 512,
   "zero_infinity": true
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f689c3b43e7accddfbf1c878981304fb35a8fda3697417b055373360d619b1f9
-size 1262054897
+oid sha256:92ce1b47197250ba9b80105b9b3a57164fd5eaa1d25043e6b5dfbc5b22589270
+size 1262058993

run.sh

@@ -13,7 +13,7 @@ python xls-r-uzbek-cv8/run_speech_recognition_ctc.py \
 	--length_column_name="input_length" \
 	--evaluation_strategy="steps" \
 	--text_column_name="sentence" \
-	--eval_metrics="wer cer" \
+	--eval_metrics="cer" \
 	--save_steps="500" \
 	--eval_steps="500" \
 	--logging_steps="100" \

run_speech_recognition_ctc.py

@@ -448,7 +448,11 @@ def main():
             batch["target_text"] = re.sub(
                 chars_to_ignore_regex, 
                 "", 
-                re.sub("([og])['`´]", "\g<1>‘", unidecode.unidecode(batch[text_column_name]).lower())
+                re.sub("['`´]", "’",   # elsewhere probably meant as glottal stop
+                       re.sub("([og])['`´]", "\g<1>‘",  # after o/g indicate modified char
+                              unidecode.unidecode(batch[text_column_name]).lower()
+                             )
+                      )
             ) + " "
         else:
             batch["target_text"] = batch[text_column_name].lower() + " "

runs/Feb02_06-54-25_job-699ba53c-fea9-4eb2-81af-a97f440eaa45/events.out.tfevents.1643785646.job-699ba53c-fea9-4eb2-81af-a97f440eaa45.33872.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bfb25063b5867cf7f135ddda4b7a60d77bbba1529b01cbea6229852aeae77d78
-size 24603
+oid sha256:3190230d80b210fe7ddda30f0da221c3e1e50ac2b1ebc18be513c0c83f125c18
+size 25074

runs/Feb02_16-57-51_job-699ba53c-fea9-4eb2-81af-a97f440eaa45/1643821174.2161925/events.out.tfevents.1643821174.job-699ba53c-fea9-4eb2-81af-a97f440eaa45.271825.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:32ca394f643571fa583c586ad18d3e1795498896953804e4db092cacc760025f
+size 4799

runs/Feb02_16-57-51_job-699ba53c-fea9-4eb2-81af-a97f440eaa45/events.out.tfevents.1643821174.job-699ba53c-fea9-4eb2-81af-a97f440eaa45.271825.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0b09a06724cb8553a167b2b4f99d617b29d0c715672c69066c343e058abb1088
+size 5852

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}, {"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}, {"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}]}
+{"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}, {"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}, {"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}]}

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:39c0625450d0afa8d2e897190721a9173256a42e1f889cdecc94feee325632c3
+oid sha256:ade06333b9174b6ec9ef767b07b6738941cd04e17f17deb11ad4726836e129b6
 size 3055

vocab.json

@@ -1 +1 @@
-{"a": 1, "b": 2, "c": 3, "d": 4, "e": 5, "f": 6, "g": 7, "h": 8, "i": 9, "j": 10, "k": 11, "l": 12, "m": 13, "n": 14, "o": 15, "p": 16, "q": 17, "r": 18, "s": 19, "t": 20, "u": 21, "v": 22, "w": 23, "x": 24, "y": 25, "z": 26, "‘": 27, "|": 0, "[UNK]": 28, "[PAD]": 29}
+{"a": 1, "b": 2, "c": 3, "d": 4, "e": 5, "f": 6, "g": 7, "h": 8, "i": 9, "j": 10, "k": 11, "l": 12, "m": 13, "n": 14, "o": 15, "p": 16, "q": 17, "r": 18, "s": 19, "t": 20, "u": 21, "v": 22, "w": 23, "x": 24, "y": 25, "z": 26, "‘": 27, "’": 28, "|": 0, "[UNK]": 29, "[PAD]": 30}