Training in progress, step 250

added_tokens.json

@@ -1 +1 @@
-{"<s>": 33, "</s>": 34}
+{"<s>": 32, "</s>": 33}

config.json

@@ -76,7 +76,7 @@
   "num_hidden_layers": 24,
   "num_negatives": 100,
   "output_hidden_size": 1024,
-  "pad_token_id": 32,
+  "pad_token_id": 31,
   "proj_codevector_dim": 768,
   "tdnn_dilation": [
     1,
@@ -102,6 +102,6 @@
   "torch_dtype": "float32",
   "transformers_version": "4.17.0.dev0",
   "use_weighted_layer_sum": false,
-  "vocab_size": 35,
+  "vocab_size": 34,
   "xvector_output_dim": 512
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2481c79419f752490f34e585207b75915a9e14fc59e3bd4b2260872480e089c7
-size 1262067185
+oid sha256:69a40f7466b01425d4ab0b61aab89e537e0e8d41f960628770413ded8e104ebc
+size 1262063089

run.sh

@@ -8,14 +8,14 @@ python run_speech_recognition_ctc.py \
         --per_device_train_batch_size="16" \
         --per_device_eval_batch_size="16" \
         --gradient_accumulation_steps="4" \
-        --learning_rate="5e-5" \
-        --warmup_steps="1000" \
+        --learning_rate="3e-5" \
+        --warmup_steps="500" \
         --length_column_name="input_length" \
         --evaluation_strategy="steps" \
         --text_column_name="text" \
-        --save_steps="500" \
-        --eval_steps="500" \
-        --logging_steps="100" \
+        --save_steps="250" \
+        --eval_steps="250" \
+        --logging_steps="50" \
         --layerdrop="0.0" \
         --activation_dropout="0.1" \
         --save_total_limit="3" \

run_speech_recognition_ctc.py

@@ -407,7 +407,7 @@ def main():
         batch["text"] = re.sub('[ç]', 'c', batch["text"])
         batch["text"] = re.sub('[úùüû]', 'u', batch["text"])
         batch["text"] = re.sub('\s', ' ', batch["text"])
-	batch["text"] = re.sub(r'\\', '', batch["text"])
+        batch["text"] = re.sub(r'\\', '', batch["text"])
         batch["text"] = re.sub('<ee>', 'eee', batch["text"])
         batch["text"] = re.sub('<qq>', 'qqq', batch["text"])
         batch["text"] = re.sub('<mm>', 'mmm', batch["text"])

runs/Jan30_14-45-48_ficino/events.out.tfevents.1643550375.ficino.242588.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c99360184fc7da7c48a6822c78596d0285442821bbd19f129f3205c3a993965e
-size 32220
+oid sha256:65756a9e9f52a0bf8493cc453294a44feec30d15e24753277bad09167cef1b35
+size 32377

runs/Jan31_10-51-54_ficino/1643622827.5155172/events.out.tfevents.1643622827.ficino.268894.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7502f1871641581204198fcad8ed3b4934654742670ef51be72521c585188815
+size 4719

runs/Jan31_10-51-54_ficino/events.out.tfevents.1643622827.ficino.268894.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7633de4c7836b94d3d45c69df74ad0d38dcdaeb9d2de5122f6294358ad5aa3ff
+size 4800

runs/Jan31_11-10-54_ficino/1643623929.218647/events.out.tfevents.1643623929.ficino.269307.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:544f6abd094e98beabe308d011fd44ddffbdcfe54f3429df6b259f6a9c9c165d
+size 4719

runs/Jan31_11-10-54_ficino/events.out.tfevents.1643623929.ficino.269307.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b13e891e3bd563eeddfa27476cf83f9370cb72c19d304183812ef827d260ea18
+size 5743

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c8e0e921ad5b5819b43e13c375604df01fb968f3613784c5df8606e8b73a0025
+oid sha256:cc6cbb10c3c21f7e4c00ecbefa42f06369abcd99e39d4879ec6f4e6804ee755e
 size 2991

vocab.json

@@ -1 +1 @@
-{"\\": 1, "a": 2, "b": 3, "c": 4, "d": 5, "e": 6, "f": 7, "g": 8, "h": 9, "i": 10, "j": 11, "k": 12, "l": 13, "m": 14, "n": 15, "o": 16, "p": 17, "q": 18, "r": 19, "s": 20, "t": 21, "u": 22, "v": 23, "w": 24, "x": 25, "y": 26, "z": 27, "å": 28, "æ": 29, "ø": 30, "|": 0, "[UNK]": 31, "[PAD]": 32}
+{"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, "å": 27, "æ": 28, "ø": 29, "|": 0, "[UNK]": 30, "[PAD]": 31}