Rolv-Arild commited on
Commit
f25415c
1 Parent(s): bb9523e

Training in progress, step 250

Browse files
.run_ex3.sh.swp ADDED
File without changes
.run_speech_recognition_ctc.py.swp ADDED
File without changes
config.json CHANGED
@@ -58,7 +58,7 @@
58
  "initializer_range": 0.02,
59
  "intermediate_size": 4096,
60
  "layer_norm_eps": 1e-05,
61
- "layerdrop": 0.0,
62
  "mask_feature_length": 64,
63
  "mask_feature_min_masks": 0,
64
  "mask_feature_prob": 0.25,
 
58
  "initializer_range": 0.02,
59
  "intermediate_size": 4096,
60
  "layer_norm_eps": 1e-05,
61
+ "layerdrop": 0.1,
62
  "mask_feature_length": 64,
63
  "mask_feature_min_masks": 0,
64
  "mask_feature_prob": 0.25,
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ab8442f42242eba6dceb2c933b909df204091bb41fcf99f2d109e6fce18f88f0
3
- size 1262063089
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bf860c4269abebda916456f37102b6c3e8e56e5f696e8c8ab73f439cba34b915
3
+ size 569727136
run.sh CHANGED
@@ -2,6 +2,7 @@ python run_speech_recognition_ctc.py \
2
  --dataset_name="NbAiLab/NPSC" \
3
  --model_name_or_path="facebook/wav2vec2-xls-r-300m" \
4
  --dataset_config_name="16K_mp3" \
 
5
  --output_dir="./" \
6
  --overwrite_output_dir \
7
  --num_train_epochs="20" \
 
2
  --dataset_name="NbAiLab/NPSC" \
3
  --model_name_or_path="facebook/wav2vec2-xls-r-300m" \
4
  --dataset_config_name="16K_mp3" \
5
+ --data_cache_dir="/mnt/lv_ai_1_ficino/rolvb/cache" \
6
  --output_dir="./" \
7
  --overwrite_output_dir \
8
  --num_train_epochs="20" \
run_ex1.sh CHANGED
@@ -2,6 +2,7 @@ python run_speech_recognition_ctc.py \
2
  --dataset_name="NbAiLab/NPSC" \
3
  --model_name_or_path="facebook/wav2vec2-xls-r-300m" \
4
  --dataset_config_name="16K_mp3" \
 
5
  --output_dir="./" \
6
  --overwrite_output_dir \
7
  --num_train_epochs="20" \
 
2
  --dataset_name="NbAiLab/NPSC" \
3
  --model_name_or_path="facebook/wav2vec2-xls-r-300m" \
4
  --dataset_config_name="16K_mp3" \
5
+ --data_cache_dir="/mnt/lv_ai_1_ficino/rolvb/cache" \
6
  --output_dir="./" \
7
  --overwrite_output_dir \
8
  --num_train_epochs="20" \
run_ex3.sh CHANGED
@@ -2,7 +2,8 @@ python run_speech_recognition_ctc.py \
2
  --dataset_name="NbAiLab/NPSC" \
3
  --model_name_or_path="facebook/wav2vec2-xls-r-300m" \
4
  --dataset_config_name="16K_mp3" \
5
- --output_dir="./ex3" \
 
6
  --overwrite_output_dir \
7
  --num_train_epochs="20" \
8
  --per_device_train_batch_size="16" \
 
2
  --dataset_name="NbAiLab/NPSC" \
3
  --model_name_or_path="facebook/wav2vec2-xls-r-300m" \
4
  --dataset_config_name="16K_mp3" \
5
+ --data_cache_dir="/mnt/lv_ai_1_ficino/rolvb/cache" \
6
+ --output_dir="./" \
7
  --overwrite_output_dir \
8
  --num_train_epochs="20" \
9
  --per_device_train_batch_size="16" \
run_speech_recognition_ctc.py CHANGED
@@ -237,6 +237,10 @@ class DataTrainingArguments:
237
  " this is only relevant if the model classifies the"
238
  " input audio to a sequence of phoneme sequences."
239
  },
 
 
 
 
240
  )
241
 
242
 
@@ -415,7 +419,7 @@ def main():
415
  if "<" in batch["text"]:
416
  raise ValueError(batch["text"])
417
  return batch
418
-
419
  # 1. First, let's load the dataset
420
  raw_datasets = DatasetDict()
421
 
@@ -425,6 +429,7 @@ def main():
425
  data_args.dataset_config_name,
426
  split=data_args.train_split_name,
427
  use_auth_token=data_args.use_auth_token,
 
428
  )
429
  raw_datasets["train"] = raw_datasets["train"].filter(filter_dataset)
430
  raw_datasets["train"] = raw_datasets["train"].map(map_dataset)
@@ -452,6 +457,7 @@ def main():
452
  data_args.dataset_config_name,
453
  split=data_args.eval_split_name,
454
  use_auth_token=data_args.use_auth_token,
 
455
  )
456
  raw_datasets["eval"] = raw_datasets["eval"].filter(filter_dataset)
457
  raw_datasets["eval"] = raw_datasets["eval"].map(map_dataset)
 
237
  " this is only relevant if the model classifies the"
238
  " input audio to a sequence of phoneme sequences."
239
  },
240
+ ),
241
+ data_cache_dir: Optional[str] = field(
242
+ default=None,
243
+ metadata={"help": "Cache directory for dataset."}
244
  )
245
 
246
 
 
419
  if "<" in batch["text"]:
420
  raise ValueError(batch["text"])
421
  return batch
422
+ assert data_args.data_cache_dir.startswith("/mnt")
423
  # 1. First, let's load the dataset
424
  raw_datasets = DatasetDict()
425
 
 
429
  data_args.dataset_config_name,
430
  split=data_args.train_split_name,
431
  use_auth_token=data_args.use_auth_token,
432
+ cache_dir=data_args.data_cache_dir
433
  )
434
  raw_datasets["train"] = raw_datasets["train"].filter(filter_dataset)
435
  raw_datasets["train"] = raw_datasets["train"].map(map_dataset)
 
457
  data_args.dataset_config_name,
458
  split=data_args.eval_split_name,
459
  use_auth_token=data_args.use_auth_token,
460
+ cache_dir=data_args.data_cache_dir
461
  )
462
  raw_datasets["eval"] = raw_datasets["eval"].filter(filter_dataset)
463
  raw_datasets["eval"] = raw_datasets["eval"].map(map_dataset)
runs/Jan31_11-10-54_ficino/events.out.tfevents.1643623929.ficino.269307.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f3b0758fcf5543564a25c3b99bc7543d4f10e275ff91404551d7c5ea2b36e4cf
3
- size 12361
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5e158a3b9ebaaebddc919c8d8108e90dd1edddd3c91466e5aafde4abe9054037
3
+ size 13146
runs/Jan31_15-12-29_ficino/1643638974.374903/events.out.tfevents.1643638974.ficino.274851.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:90da1e9403bc415312a147da7992baa0d63d67df4e5f38ccefe70208b63de90f
3
+ size 4719
runs/Jan31_15-12-29_ficino/events.out.tfevents.1643638974.ficino.274851.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e4f2ca8390f2c1035d8a2d0991da63d85ca0b8c13ed1994b0f914c2346cfddb
3
+ size 5745
runs/Jan31_15-14-01_ficino/1643638993.9934366/events.out.tfevents.1643638993.ficino.275087.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:01f0849ee2681cec8dc1acc83de8ba27458f7b861b504892463652e8c54b3260
3
+ size 4719
runs/Jan31_15-14-01_ficino/events.out.tfevents.1643638993.ficino.275087.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:002e37168a9da5b8b0a35fef88b252c405be26248cae2b71357cea32785d8df0
3
+ size 5743
special_tokens_map.json CHANGED
@@ -1 +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}]}
 
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}]}
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:cc6cbb10c3c21f7e4c00ecbefa42f06369abcd99e39d4879ec6f4e6804ee755e
3
  size 2991
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8a47239e05c536ae72f6dfb5b0171a4189827226ec67e1a549e250529ffefac8
3
  size 2991