pere commited on Mar 17, 2023

Commit

1189f37

•

1 Parent(s): e707816

Training in progress, step 200

Files changed (18) hide show

.gitignore ADDED Viewed

	@@ -0,0 +1 @@


1	+ checkpoint-*/

added_tokens.json CHANGED Viewed

@@ -17,7 +17,6 @@
   "<|da|>": 50285,
   "<|de|>": 50261,
   "<|el|>": 50281,
-  "<|endoftext|>": 50257,
   "<|en|>": 50259,
   "<|es|>": 50262,
   "<|et|>": 50307,
@@ -30,6 +29,7 @@
   "<|gu|>": 50333,
   "<|haw|>": 50352,
   "<|ha|>": 50354,
   "<|hi|>": 50276,
   "<|hr|>": 50291,
   "<|ht|>": 50339,
@@ -38,7 +38,6 @@
   "<|id|>": 50275,
   "<|is|>": 50311,
   "<|it|>": 50274,
-  "<|iw|>": 50279,
   "<|ja|>": 50266,
   "<|jw|>": 50356,
   "<|ka|>": 50329,

   "<|da|>": 50285,
   "<|de|>": 50261,
   "<|el|>": 50281,
   "<|en|>": 50259,
   "<|es|>": 50262,
   "<|et|>": 50307,
   "<|gu|>": 50333,
   "<|haw|>": 50352,
   "<|ha|>": 50354,
+  "<|he|>": 50279,
   "<|hi|>": 50276,
   "<|hr|>": 50291,
   "<|ht|>": 50339,
   "<|id|>": 50275,
   "<|is|>": 50311,
   "<|it|>": 50274,
   "<|ja|>": 50266,
   "<|jw|>": 50356,
   "<|ka|>": 50329,

config.json CHANGED Viewed

@@ -2,6 +2,7 @@
   "_name_or_path": "openai/whisper-tiny",
   "activation_dropout": 0.0,
   "activation_function": "gelu",
   "architectures": [
     "WhisperForConditionalGeneration"
   ],
@@ -11,6 +12,7 @@
     50257
   ],
   "bos_token_id": 50257,
   "d_model": 384,
   "decoder_attention_heads": 6,
   "decoder_ffn_dim": 1536,
@@ -26,6 +28,12 @@
   "forced_decoder_ids": null,
   "init_std": 0.02,
   "is_encoder_decoder": true,
   "max_length": 448,
   "max_source_positions": 1500,
   "max_target_positions": 448,
@@ -35,7 +43,8 @@
   "pad_token_id": 50257,
   "scale_embedding": false,
   "torch_dtype": "float32",
-  "transformers_version": "4.26.0.dev0",
   "use_cache": false,
   "vocab_size": 51865
 }

   "_name_or_path": "openai/whisper-tiny",
   "activation_dropout": 0.0,
   "activation_function": "gelu",
+  "apply_spec_augment": false,
   "architectures": [
     "WhisperForConditionalGeneration"
   ],
     50257
   ],
   "bos_token_id": 50257,
+  "classifier_proj_size": 256,
   "d_model": 384,
   "decoder_attention_heads": 6,
   "decoder_ffn_dim": 1536,
   "forced_decoder_ids": null,
   "init_std": 0.02,
   "is_encoder_decoder": true,
+  "mask_feature_length": 10,
+  "mask_feature_min_masks": 0,
+  "mask_feature_prob": 0.0,
+  "mask_time_length": 10,
+  "mask_time_min_masks": 2,
+  "mask_time_prob": 0.05,
   "max_length": 448,
   "max_source_positions": 1500,
   "max_target_positions": 448,
   "pad_token_id": 50257,
   "scale_embedding": false,
   "torch_dtype": "float32",
+  "transformers_version": "4.28.0.dev0",
   "use_cache": false,
+  "use_weighted_layer_sum": false,
   "vocab_size": 51865
 }

merges.txt CHANGED Viewed

@@ -1,4 +1,4 @@
-#version: 0.2
 Ġ a
 Ġt h
 i n

+#version: 0.2 - Trained by `huggingface/tokenizers`
 Ġ a
 Ġt h
 i n

preprocessor_config.json CHANGED Viewed

The diff for this file is too large to render. See raw diff

pytorch_model.bin ADDED Viewed

+version https://git-lfs.github.com/spec/v1
+oid sha256:12553e0e927eed2c74d99b6cf8b63f39c52c3d7ddf6685b691ef4eaf737aa78b
+size 151098921

run.sh CHANGED Viewed

@@ -9,7 +9,6 @@ python run_speech_recognition_seq2seq_streaming.py \
 	--output_dir="./" \
 	--per_device_train_batch_size="128" \
 	--per_device_eval_batch_size="32" \
-	--gradient_accumulation_step="1" \
 	--logging_steps="200" \
 	--learning_rate="3e-6" \
 	--lr_scheduler_type="constant_with_warmup" \

 	--output_dir="./" \
 	--per_device_train_batch_size="128" \
 	--per_device_eval_batch_size="32" \
 	--logging_steps="200" \
 	--learning_rate="3e-6" \
 	--lr_scheduler_type="constant_with_warmup" \

runs/Mar17_11-26-25_dante/1679048805.5259922/events.out.tfevents.1679048805.dante.3289286.1 ADDED Viewed

+version https://git-lfs.github.com/spec/v1
+oid sha256:eb323c0adaf46f08789eaa656ce29f54a414144a4d7089e6e5e6b6b78d816386
+size 6000

runs/Mar17_11-26-25_dante/events.out.tfevents.1679048805.dante.3289286.0 ADDED Viewed

+version https://git-lfs.github.com/spec/v1
+oid sha256:96111a5cc0a61b189c72c37afde9e60957029a82678ef276b338089b3ad5975d
+size 4657

runs/Mar17_11-44-55_dante/1679049913.0870838/events.out.tfevents.1679049913.dante.3379179.1 ADDED Viewed

+version https://git-lfs.github.com/spec/v1
+oid sha256:dad130fea0ce6160e7e3f977432f0dbe443d9e7b5528bc3186c98b110a5a70d4
+size 6000

runs/Mar17_11-44-55_dante/events.out.tfevents.1679049913.dante.3379179.0 ADDED Viewed

+version https://git-lfs.github.com/spec/v1
+oid sha256:0c903331b7bf4b5917942facf549c8fd588774ef29d1c1eb5ff53e2c52c70337
+size 4136

runs/Mar17_11-46-06_dante/1679049983.539156/events.out.tfevents.1679049983.dante.3389725.1 ADDED Viewed

+version https://git-lfs.github.com/spec/v1
+oid sha256:529285418f7d91b1c6a78c8472f127c1d56cde29c73e72b738f8f57183c2a5bf
+size 6000

runs/Mar17_11-46-06_dante/events.out.tfevents.1679049983.dante.3389725.0 ADDED Viewed

+version https://git-lfs.github.com/spec/v1
+oid sha256:f68fc2531410ac9a6101deeb5352ac6cd5136feb1657011faea65415aa401bab
+size 5132

special_tokens_map.json CHANGED Viewed

@@ -22,7 +22,7 @@
     "<|hi|>",
     "<|fi|>",
     "<|vi|>",
-    "<|iw|>",
     "<|uk|>",
     "<|el|>",
     "<|ms|>",
@@ -124,7 +124,7 @@
   },
   "pad_token": "<|endoftext|>",
   "unk_token": {
-    "content": "",
     "lstrip": false,
     "normalized": true,
     "rstrip": false,

     "<|hi|>",
     "<|fi|>",
     "<|vi|>",
+    "<|he|>",
     "<|uk|>",
     "<|el|>",
     "<|ms|>",
   },
   "pad_token": "<|endoftext|>",
   "unk_token": {
+    "content": "<|endoftext|>",
     "lstrip": false,
     "normalized": true,
     "rstrip": false,

tokenizer.json ADDED Viewed

The diff for this file is too large to render. See raw diff

tokenizer_config.json CHANGED Viewed

@@ -19,7 +19,6 @@
   },
   "errors": "replace",
   "model_max_length": 1024,
-  "name_or_path": "openai/whisper-tiny",
   "pad_token": null,
   "processor_class": "WhisperProcessor",
   "return_attention_mask": false,
@@ -27,7 +26,7 @@
   "tokenizer_class": "WhisperTokenizer",
   "unk_token": {
     "__type": "AddedToken",
-    "content": "",
     "lstrip": false,
     "normalized": true,
     "rstrip": false,

   },
   "errors": "replace",
   "model_max_length": 1024,
   "pad_token": null,
   "processor_class": "WhisperProcessor",
   "return_attention_mask": false,
   "tokenizer_class": "WhisperTokenizer",
   "unk_token": {
     "__type": "AddedToken",
+    "content": "<|endoftext|>",
     "lstrip": false,
     "normalized": true,
     "rstrip": false,

training_args.bin CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:402de2c4bd980e7aa9dfbd8a32ec503e2002f70e226ac0ad0b28d4f63153cc68
-size 3643

 version https://git-lfs.github.com/spec/v1
+oid sha256:2e571ee67bac8da76cc9d321b972299d9ace7203f12f58e3e1067698c4cb42d0
+size 3707

vocab.json CHANGED Viewed

The diff for this file is too large to render. See raw diff