Initial commit

.gitattributes

@@ -0,0 +1,16 @@
+*.bin.* filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tar.gz filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,181 @@
+---
+language: zh
+datasets:
+- common_voice
+tags:
+- speech
+- audio
+- automatic-speech-recognition
+- xlsr_fine_tuning_week
+license: apache-2.0
+---
+
+## Colab trial with recording or voice file
+[Colab trial](https://colab.research.google.com/drive/1e_z5jQHYbO2YKEaUgzb1ww1WwiAyydAj?usp=sharing)
+
+```
+import torchaudio
+from datasets import load_dataset, load_metric
+from transformers import (
+    Wav2Vec2ForCTC,
+    Wav2Vec2Processor,
+)
+import torch
+import re
+import sys
+
+model_name = "voidful/wav2vec2-large-xlsr-53-tw"
+device = "cuda"
+processor_name = "voidful/wav2vec2-large-xlsr-53-tw"
+
+chars_to_ignore_regex = r"[¥•＂＃＄％＆＇（）＊＋，－／：；＜＝＞＠［＼］＾＿｀｛｜｝～｟｠｢｣､　、〃〈〉《》「」『』【】〔〕〖〗〘〙〚〛〜〝〞〟〰〾〿–—‘’‛“”„‟…‧﹏﹑﹔·'℃°•·．﹑︰〈〉─《﹖﹣﹂﹁﹔！？｡。＂＃＄％＆＇（）＊＋，﹐－／：；＜＝＞＠［＼］＾＿｀｛｜｝～｟｠｢｣､、〃》「」『』【】〔〕〖〗〘〙〚〛〜〝〞〟〰〾〿–—‘’‛“”„‟…‧﹏.．!\"#$%&()*+,\-.\:;<=>?@\[\]\\\/^_`{|}~]"
+
+model = Wav2Vec2ForCTC.from_pretrained(model_name).to(device)
+processor = Wav2Vec2Processor.from_pretrained(processor_name)
+
+resampler = torchaudio.transforms.Resample(orig_freq=48_000, new_freq=16_000)
+
+def load_file_to_data(file):
+    batch = {}
+    speech, _ = torchaudio.load(file)
+    batch["speech"] = resampler.forward(speech.squeeze(0)).numpy()
+    batch["sampling_rate"] = resampler.new_freq
+    return batch
+
+
+def predict(data):
+    features = processor(data["speech"], sampling_rate=data["sampling_rate"], padding=True, return_tensors="pt")
+    input_values = features.input_values.to(device)
+    attention_mask = features.attention_mask.to(device)
+    with torch.no_grad():
+        logits = model(input_values, attention_mask=attention_mask).logits
+    pred_ids = torch.argmax(logits, dim=-1)
+    return processor.batch_decode(pred_ids)
+    
+```
+
+Predict
+```python
+predict(load_file_to_data('voice file path'))
+```
+
+## Evaluation on Common Voice TW Test
+```python
+import torchaudio
+from datasets import load_dataset, load_metric
+from transformers import (
+    Wav2Vec2ForCTC,
+    Wav2Vec2Processor,
+)
+import torch
+import re
+
+model_name = "voidful/wav2vec2-large-xlsr-53-tw"
+device = "cuda"
+processor_name = "voidful/wav2vec2-large-xlsr-53-tw"
+
+chars_to_ignore_regex = r"[¥•＂＃＄％＆＇（）＊＋，－／：；＜＝＞＠［＼］＾＿｀｛｜｝～｟｠｢｣､　、〃〈〉《》「」『』【】〔〕〖〗〘〙〚〛〜〝〞〟〰〾〿–—‘’‛“”„‟…‧﹏﹑﹔·'℃°•·．﹑︰〈〉─《﹖﹣﹂﹁﹔！？｡。＂＃＄％＆＇（）＊＋，﹐－／：；＜＝＞＠［＼］＾＿｀｛｜｝～｟｠｢｣､、〃》「」『』【】〔〕〖〗〘〙〚〛〜〝〞〟〰〾〿–—‘’‛“”„‟…‧﹏.．!\"#$%&()*+,\-.\:;<=>?@\[\]\\\/^_`{|}~]"
+
+model = Wav2Vec2ForCTC.from_pretrained(model_name).to(device)
+processor = Wav2Vec2Processor.from_pretrained(processor_name)
+
+ds = load_dataset("common_voice", 'zh-TW', data_dir="./cv-corpus-6.1-2020-12-11", split="test")
+
+resampler = torchaudio.transforms.Resample(orig_freq=48_000, new_freq=16_000)
+
+def map_to_array(batch):
+    speech, _ = torchaudio.load(batch["path"])
+    batch["speech"] = resampler.forward(speech.squeeze(0)).numpy()
+    batch["sampling_rate"] = resampler.new_freq
+    batch["sentence"] = re.sub(chars_to_ignore_regex, '', batch["sentence"]).lower().replace("’", "'")
+    return batch
+
+ds = ds.map(map_to_array)
+
+def map_to_pred(batch):
+    features = processor(batch["speech"], sampling_rate=batch["sampling_rate"][0], padding=True, return_tensors="pt")
+    input_values = features.input_values.to(device)
+    attention_mask = features.attention_mask.to(device)
+    with torch.no_grad():
+        logits = model(input_values, attention_mask=attention_mask).logits
+    pred_ids = torch.argmax(logits, dim=-1)
+    batch["predicted"] = processor.batch_decode(pred_ids)
+    batch["target"] = batch["sentence"]
+    return batch
+
+result = ds.map(map_to_pred, batched=True, batch_size=16, remove_columns=list(ds.features.keys()))
+
+wer = load_metric("wer")
+
+print(wer.compute(predictions=result["predicted"], references=result["target"]))
+```
+
+`CER: 0.8635578583765112`
+
+Inference with GPT LM:
+```python
+import torchaudio
+from datasets import load_dataset, load_metric
+from transformers import (
+    Wav2Vec2ForCTC,
+    Wav2Vec2Processor,
+)
+import torch
+import re
+from transformers import AutoTokenizer, AutoModelWithLMHead 
+
+model_name = "voidful/wav2vec2-large-xlsr-53-tw"
+device = "cuda"
+processor_name = "voidful/wav2vec2-large-xlsr-53-tw"
+
+chars_to_ignore_regex = r"[¥•＂＃＄％＆＇（）＊＋，－／：；＜＝＞＠［＼］＾＿｀｛｜｝～｟｠｢｣､　、〃〈〉《》「」『』【】〔〕〖〗〘〙〚〛〜〝〞���〰〾〿–—‘’‛“”„‟…‧﹏﹑﹔·'℃°•·．﹑︰〈〉─《﹖﹣﹂﹁﹔！？｡。＂＃＄％＆＇（）＊＋，﹐－／：；＜＝＞＠［＼］＾＿｀｛｜｝～｟｠｢｣､、〃》「」『』【】〔〕〖〗〘〙〚〛〜〝〞〟〰〾〿–—‘’‛“”„‟…‧﹏.．!\"#$%&()*+,\-.\:;<=>?@\[\]\\\/^_`{|}~]"
+
+tokenizer = AutoTokenizer.from_pretrained("ckiplab/gpt2-base-chinese")  
+gpt_model = AutoModelWithLMHead.from_pretrained("ckiplab/gpt2-base-chinese").to(device)
+model = Wav2Vec2ForCTC.from_pretrained(model_name).to(device)
+processor = Wav2Vec2Processor.from_pretrained(processor_name)
+
+ds = load_dataset("common_voice", 'zh-TW', data_dir="./cv-corpus-6.1-2020-12-11", split="test")
+
+resampler = torchaudio.transforms.Resample(orig_freq=48_000, new_freq=16_000)
+
+def map_to_array(batch):
+    speech, _ = torchaudio.load(batch["path"])
+    batch["speech"] = resampler.forward(speech.squeeze(0)).numpy()
+    batch["sampling_rate"] = resampler.new_freq
+    batch["sentence"] = re.sub(chars_to_ignore_regex, '', batch["sentence"]).lower().replace("’", "'")
+    return batch
+
+ds = ds.map(map_to_array)
+
+def map_to_pred(batch):
+    features = processor(batch["speech"], sampling_rate=batch["sampling_rate"][0], padding=True, return_tensors="pt")
+    input_values = features.input_values.to(device)
+    attention_mask = features.attention_mask.to(device)
+    with torch.no_grad():
+        logits = model(input_values, attention_mask=attention_mask).logits
+    
+    decoded_results = []
+    for logit in logits:
+        pred_ids = torch.argmax(logit, dim=-1)
+        mask = pred_ids.ge(1).unsqueeze(-1).expand(logit.size())
+        vocab_size = logit.size()[-1]
+        voice_prob = torch.nn.functional.softmax((torch.masked_select(logit, mask).view(-1,vocab_size)),dim=-1)
+        gpt_input = torch.cat((torch.tensor([tokenizer.cls_token_id]).to(device),pred_ids[pred_ids>0]), 0)
+        gpt_prob = torch.nn.functional.softmax(gpt_model(gpt_input).logits, dim=-1)[:voice_prob.size()[0],:]
+        comb_pred_ids = torch.argmax(gpt_prob*voice_prob, dim=-1)
+        decoded_results.append(processor.decode(comb_pred_ids))
+        
+    batch["predicted"] = decoded_results
+    batch["target"] = batch["sentence"]
+    return batch
+
+
+result = ds.map(map_to_pred, batched=True, batch_size=16, remove_columns=list(ds.features.keys()))
+
+wer = load_metric("wer")
+
+print(wer.compute(predictions=result["predicted"], references=result["target"]))
+```
+
+`CER 0.7927461139896373`

config.json

@@ -0,0 +1,76 @@
+{
+  "_name_or_path": "facebook/wav2vec2-large-xlsr-53",
+  "activation_dropout": 0.0,
+  "apply_spec_augment": true,
+  "architectures": [
+    "Wav2Vec2ForCTC"
+  ],
+  "attention_dropout": 0.1,
+  "bos_token_id": 1,
+  "conv_bias": true,
+  "conv_dim": [
+    512,
+    512,
+    512,
+    512,
+    512,
+    512,
+    512
+  ],
+  "conv_kernel": [
+    10,
+    3,
+    3,
+    3,
+    3,
+    2,
+    2
+  ],
+  "conv_stride": [
+    5,
+    2,
+    2,
+    2,
+    2,
+    2,
+    2
+  ],
+  "ctc_loss_reduction": "mean",
+  "ctc_zero_infinity": false,
+  "do_stable_layer_norm": true,
+  "eos_token_id": 2,
+  "feat_extract_activation": "gelu",
+  "feat_extract_dropout": 0.0,
+  "feat_extract_norm": "layer",
+  "feat_proj_dropout": 0.0,
+  "final_dropout": 0.0,
+  "gradient_checkpointing": true,
+  "hidden_act": "gelu",
+  "hidden_dropout": 0.1,
+  "hidden_size": 1024,
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "layer_norm_eps": 1e-05,
+  "layerdrop": 0.1,
+  "mask_channel_length": 10,
+  "mask_channel_min_space": 1,
+  "mask_channel_other": 0.0,
+  "mask_channel_prob": 0.0,
+  "mask_channel_selection": "static",
+  "mask_feature_length": 10,
+  "mask_feature_prob": 0.0,
+  "mask_time_length": 10,
+  "mask_time_min_space": 1,
+  "mask_time_other": 0.0,
+  "mask_time_prob": 0.05,
+  "mask_time_selection": "static",
+  "model_type": "wav2vec2",
+  "num_attention_heads": 16,
+  "num_conv_pos_embedding_groups": 16,
+  "num_conv_pos_embeddings": 128,
+  "num_feat_extract_layers": 7,
+  "num_hidden_layers": 24,
+  "pad_token_id": 0,
+  "transformers_version": "4.4.0.dev0",
+  "vocab_size": 21128
+}

feature_extractor_config.json

@@ -0,0 +1,9 @@
+{
+    "do_normalize": true,
+    "feature_size": 1,
+    "padding_side": "right",
+    "padding_value": 0.0,
+    "return_attention_mask": true,
+    "sampling_rate": 16000
+  }
+  

preprocessor_config.json

@@ -0,0 +1,8 @@
+{
+  "do_normalize": true,
+  "feature_size": 1,
+  "padding_side": "right",
+  "padding_value": 0.0,
+  "return_attention_mask": true,
+  "sampling_rate": 16000
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:80b6393d293991d01b78ee9219ad2a32d65cb30f33433491a795de7879322b8a
+size 1348554373

special_tokens_map.json

@@ -0,0 +1 @@
+{"unk_token": "[UNK]", "pad_token": "[PAD]"}

tokenizer_config.json

@@ -0,0 +1 @@
+{"unk_token": "[UNK]", "bos_token": null, "eos_token": null, "pad_token": "[PAD]", "do_lower_case": false, "word_delimiter_token": "|"}