Wav2Vec Malayalam with Ngrams

.gitattributes

@@ -1,35 +1,18 @@
-*.7z filter=lfs diff=lfs merge=lfs -text
-*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin.* filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
 *.bin filter=lfs diff=lfs merge=lfs -text
-*.bz2 filter=lfs diff=lfs merge=lfs -text
-*.ckpt filter=lfs diff=lfs merge=lfs -text
-*.ftz filter=lfs diff=lfs merge=lfs -text
-*.gz 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
-*.lfs.* filter=lfs diff=lfs merge=lfs -text
-*.mlmodel filter=lfs diff=lfs merge=lfs -text
 *.model filter=lfs diff=lfs merge=lfs -text
 *.msgpack filter=lfs diff=lfs merge=lfs -text
-*.npy filter=lfs diff=lfs merge=lfs -text
-*.npz filter=lfs diff=lfs merge=lfs -text
-*.onnx filter=lfs diff=lfs merge=lfs -text
-*.ot filter=lfs diff=lfs merge=lfs -text
-*.parquet filter=lfs diff=lfs merge=lfs -text
 *.pb filter=lfs diff=lfs merge=lfs -text
-*.pickle filter=lfs diff=lfs merge=lfs -text
-*.pkl filter=lfs diff=lfs merge=lfs -text
 *.pt filter=lfs diff=lfs merge=lfs -text
 *.pth filter=lfs diff=lfs merge=lfs -text
-*.rar filter=lfs diff=lfs merge=lfs -text
-*.safetensors filter=lfs diff=lfs merge=lfs -text
-saved_model/**/* filter=lfs diff=lfs merge=lfs -text
-*.tar.* filter=lfs diff=lfs merge=lfs -text
-*.tar filter=lfs diff=lfs merge=lfs -text
-*.tflite filter=lfs diff=lfs merge=lfs -text
-*.tgz filter=lfs diff=lfs merge=lfs -text
-*.wasm filter=lfs diff=lfs merge=lfs -text
-*.xz filter=lfs diff=lfs merge=lfs -text
-*.zip filter=lfs diff=lfs merge=lfs -text
-*.zst filter=lfs diff=lfs merge=lfs -text
-*tfevents* filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+language_model/5gram_correct.arpa filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,182 @@
+---
+language: ml
+datasets:
+- Indic TTS Malayalam Speech Corpus
+- Openslr Malayalam Speech Corpus
+- SMC Malayalam Speech Corpus
+- IIIT-H Indic Speech Databases
+metrics:
+- wer
+tags:
+- audio
+- automatic-speech-recognition
+- speech
+- xlsr-fine-tuning-week
+license: apache-2.0
+model-index:
+- name: Malayalam XLSR Wav2Vec2 Large 53
+  results:
+  - task: 
+      name: Speech Recognition
+      type: automatic-speech-recognition
+    dataset:
+      name: Test split of combined dataset using all datasets mentioned above
+      type: custom
+      args: ml
+    metrics:
+       - name: Test WER
+         type: wer
+         value: 28.43
+---
+
+# Wav2Vec2-Large-XLSR-53-ml
+
+Fine-tuned [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on ml (Malayalam) using the [Indic TTS Malayalam Speech Corpus (via Kaggle)](https://www.kaggle.com/kavyamanohar/indic-tts-malayalam-speech-corpus), [Openslr Malayalam Speech Corpus](http://openslr.org/63/), [SMC Malayalam Speech Corpus](https://blog.smc.org.in/malayalam-speech-corpus/) and [IIIT-H Indic Speech Databases](http://speech.iiit.ac.in/index.php/research-svl/69.html). The notebooks used to train model are available [here](https://github.com/gauthamsuresh09/wav2vec2-large-xlsr-53-malayalam/). When using this model, make sure that your speech input is sampled at 16kHz.
+
+## Usage
+
+The model can be used directly (without a language model) as follows:
+
+```python
+import torch
+import torchaudio
+from datasets import load_dataset
+from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
+
+test_dataset = <load-test-split-of-combined-dataset> # Details on loading this dataset in the evaluation section
+
+processor = Wav2Vec2Processor.from_pretrained("gvs/wav2vec2-large-xlsr-malayalam")
+model = Wav2Vec2ForCTC.from_pretrained("gvs/wav2vec2-large-xlsr-malayalam")
+
+resampler = torchaudio.transforms.Resample(48_000, 16_000)
+
+# Preprocessing the datasets.
+# We need to read the audio files as arrays
+def speech_file_to_array_fn(batch):
+  speech_array, sampling_rate = torchaudio.load(batch["path"])
+  batch["speech"] = resampler(speech_array).squeeze().numpy()
+  return batch
+
+test_dataset = test_dataset.map(speech_file_to_array_fn)
+inputs = processor(test_dataset["speech"][:2], sampling_rate=16_000, return_tensors="pt", padding=True)
+
+with torch.no_grad():
+  logits = model(inputs.input_values, attention_mask=inputs.attention_mask).logits
+
+predicted_ids = torch.argmax(logits, dim=-1)
+
+print("Prediction:", processor.batch_decode(predicted_ids))
+print("Reference:", test_dataset["sentence"])
+```
+
+
+## Evaluation
+
+The model can be evaluated as follows on the test data of combined custom dataset. For more details on dataset preparation, check the notebooks mentioned at the end of this file.
+
+
+```python
+import torch
+import torchaudio
+from datasets import load_dataset, load_metric
+from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
+import re
+from datasets import load_dataset, load_metric
+from pathlib import Path
+
+# The custom dataset needs to be created using notebook mentioned at the end of this file
+data_dir = Path('<path-to-custom-dataset>')
+
+dataset_folders = {
+    'iiit': 'iiit_mal_abi',
+    'openslr': 'openslr',
+    'indic-tts': 'indic-tts-ml',
+    'msc-reviewed': 'msc-reviewed-speech-v1.0+20200825',
+}
+
+# Set directories for datasets
+openslr_male_dir = data_dir / dataset_folders['openslr'] / 'male'
+openslr_female_dir = data_dir / dataset_folders['openslr'] / 'female'
+iiit_dir = data_dir / dataset_folders['iiit']
+indic_tts_male_dir = data_dir / dataset_folders['indic-tts'] / 'male'
+indic_tts_female_dir = data_dir / dataset_folders['indic-tts'] / 'female'
+msc_reviewed_dir = data_dir / dataset_folders['msc-reviewed']
+
+# Load the datasets
+openslr_male = load_dataset("json", data_files=[f"{str(openslr_male_dir.absolute())}/sample_{i}.json" for i in range(2023)], split="train")
+openslr_female = load_dataset("json", data_files=[f"{str(openslr_female_dir.absolute())}/sample_{i}.json" for i in range(2103)], split="train")
+iiit = load_dataset("json", data_files=[f"{str(iiit_dir.absolute())}/sample_{i}.json" for i in range(1000)], split="train")
+indic_tts_male = load_dataset("json", data_files=[f"{str(indic_tts_male_dir.absolute())}/sample_{i}.json" for i in range(5649)], split="train")
+indic_tts_female = load_dataset("json", data_files=[f"{str(indic_tts_female_dir.absolute())}/sample_{i}.json" for i in range(2950)], split="train")
+msc_reviewed = load_dataset("json", data_files=[f"{str(msc_reviewed_dir.absolute())}/sample_{i}.json" for i in range(1541)], split="train")
+
+# Create test split as 20%, set random seed as well.
+test_size = 0.2
+random_seed=1
+openslr_male_splits = openslr_male.train_test_split(test_size=test_size, seed=random_seed)
+openslr_female_splits = openslr_female.train_test_split(test_size=test_size, seed=random_seed)
+iiit_splits = iiit.train_test_split(test_size=test_size, seed=random_seed)
+indic_tts_male_splits = indic_tts_male.train_test_split(test_size=test_size, seed=random_seed)
+indic_tts_female_splits = indic_tts_female.train_test_split(test_size=test_size, seed=random_seed)
+msc_reviewed_splits = msc_reviewed.train_test_split(test_size=test_size, seed=random_seed)
+
+# Get combined test dataset
+split_list = [openslr_male_splits, openslr_female_splits, indic_tts_male_splits, indic_tts_female_splits, msc_reviewed_splits, iiit_splits]
+test_dataset = datasets.concatenate_datasets([split['test'] for split in split_list)
+
+wer = load_metric("wer")
+
+processor = Wav2Vec2Processor.from_pretrained("gvs/wav2vec2-large-xlsr-malayalam") 
+model = Wav2Vec2ForCTC.from_pretrained("gvs/wav2vec2-large-xlsr-malayalam")
+model.to("cuda")
+
+resamplers = {
+    48000: torchaudio.transforms.Resample(48_000, 16_000),
+}
+
+chars_to_ignore_regex = '[\\\\,\\\\?\\\\.\\\\!\\\\-\\\\;\\\\:\\\\"\\\\“\\\\%\\\\‘\\\\”\\\\�Utrnle\\\\_]'
+unicode_ignore_regex = r'[\\\\u200e]'
+
+# Preprocessing the datasets.
+# We need to read the audio files as arrays
+def speech_file_to_array_fn(batch):
+    batch["sentence"] = re.sub(chars_to_ignore_regex, '', batch["sentence"])
+    batch["sentence"] = re.sub(unicode_ignore_regex, '', batch["sentence"])
+    speech_array, sampling_rate = torchaudio.load(batch["path"])
+    # Resample if its not in 16kHz
+    if sampling_rate != 16000:
+        batch["speech"] = resamplers[sampling_rate](speech_array).squeeze().numpy()
+    else:
+        batch["speech"] = speech_array.squeeze().numpy()
+    # If more than one dimension is present, pick first one
+    if batch["speech"].ndim > 1:
+        batch["speech"] = batch["speech"][0]
+    return batch
+
+test_dataset = test_dataset.map(speech_file_to_array_fn)
+
+# Preprocessing the datasets.
+# We need to read the audio files as arrays
+def evaluate(batch):
+  inputs = processor(batch["speech"], sampling_rate=16_000, return_tensors="pt", padding=True)
+
+  with torch.no_grad():
+    logits = model(inputs.input_values.to("cuda"), attention_mask=inputs.attention_mask.to("cuda")).logits
+
+  pred_ids = torch.argmax(logits, dim=-1)
+  batch["pred_strings"] = processor.batch_decode(pred_ids)
+  return batch
+
+result = test_dataset.map(evaluate, batched=True, batch_size=8)
+
+print("WER: {:2f}".format(100 * wer.compute(predictions=result["pred_strings"], references=result["sentence"])))
+```
+
+**Test Result (WER)**: 28.43 %  
+
+
+## Training
+
+A combined dataset was created using [Indic TTS Malayalam Speech Corpus (via Kaggle)](https://www.kaggle.com/kavyamanohar/indic-tts-malayalam-speech-corpus), [Openslr Malayalam Speech Corpus](http://openslr.org/63/), [SMC Malayalam Speech Corpus](https://blog.smc.org.in/malayalam-speech-corpus/) and [IIIT-H Indic Speech Databases](http://speech.iiit.ac.in/index.php/research-svl/69.html). The datasets were downloaded and was converted to HF Dataset format using [this notebook](https://github.com/gauthamsuresh09/wav2vec2-large-xlsr-53-malayalam/blob/main/make_hf_dataset.ipynb)
+
+The notebook used for training and evaluation can be found [here](https://github.com/gauthamsuresh09/wav2vec2-large-xlsr-53-malayalam/blob/main/fine-tune-xlsr-wav2vec2-on-malayalam-asr-with-transformers_v2.ipynb)

added_tokens.json

@@ -0,0 +1,4 @@
+{
+  "</s>": 77,
+  "<s>": 76
+}

alphabet.json

@@ -0,0 +1 @@
+{"labels": ["\u0d7c", "\u0d38", "\u0d02", "\u0d0b", "\u0d39", "\u0d27", "\u0d22", "\u0d20", "\u0d1f", "\u0d2a", "\u0d2f", "\u0d40", "\u0d15", " ", "\u0d30", "\u0d0f", "\u0d09", "\u0d43", "\u0d7a", "\u0d23", "\u0d3f", "\u200c", "\u0d7d", "\u0d4a", "\u0d32", "\u0d57", "\u0d18", "\u0d4c", "\u0d26", "\u0d2c", "\u0d1b", "\u0d1d", "\u0d0a", "\u0d28", "\u0d4b", "\u0d71", "\u0d07", "\u0d10", "\u0d48", "\u0d03", "\u0d3e", "\u0d1a", "\u0d47", "\u0d25", "\u0d21", "\u0d0e", "\u0d37", "\u0d06", "\u0d05", "\u0d2e", "\u0d12", "\u0d36", "\u0d16", "\u0d2b", "\u0d41", "\u0d17", "\u0d33", "\u0d13", "\u0d42", "\u0d35", "\u0d1e", "\u0d31", "\u0d2d", "\u0d7b", "\u0d7e", "\u0d08", "\u0d4d", "\u0d46", "\u0d19", "\u0d24", "\u0d1c", "\u200d", "\u0d14", "\u0d34", "\u2047", "", "<s>", "</s>"], "is_bpe": false}

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": 0.094,
+  "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": 0.05,
+  "hidden_size": 1024,
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "layer_norm_eps": 1e-05,
+  "layerdrop": 0.045,
+  "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": 0.082,
+  "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": 75,
+  "transformers_version": "4.5.0.dev0",
+  "vocab_size": 76
+}

flax_model.msgpack

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:18381d01f5140589dfb1d13754725c976fd7172be4e92a52445f2a13baf638f6
+size 1262081874

language_model/5gram_correct.arpa

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b0bb3d117437dfa2111ac7f68eb4e2a411532515a3c10cb7abff205455ec8852
+size 15359294

language_model/attrs.json

@@ -0,0 +1 @@
+{"alpha": 0.5, "beta": 1.5, "unk_score_offset": -10.0, "score_boundary": true}

preprocessor_config.json

@@ -0,0 +1,10 @@
+{
+  "do_normalize": true,
+  "feature_extractor_type": "Wav2Vec2FeatureExtractor",
+  "feature_size": 1,
+  "padding_side": "right",
+  "padding_value": 0.0,
+  "processor_class": "Wav2Vec2ProcessorWithLM",
+  "return_attention_mask": true,
+  "sampling_rate": 16000
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5c39014b08099f35aa8eab263323ff24c61f959eddd9d50dc9716ab0df8afee
+size 1262245399

special_tokens_map.json

@@ -0,0 +1,6 @@
+{
+  "bos_token": "<s>",
+  "eos_token": "</s>",
+  "pad_token": "[PAD]",
+  "unk_token": "[UNK]"
+}

tokenizer_config.json

@@ -0,0 +1,48 @@
+{
+  "added_tokens_decoder": {
+    "74": {
+      "content": "[UNK]",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": false
+    },
+    "75": {
+      "content": "[PAD]",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": false
+    },
+    "76": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "77": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": true,
+  "do_lower_case": false,
+  "eos_token": "</s>",
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "[PAD]",
+  "processor_class": "Wav2Vec2ProcessorWithLM",
+  "replace_word_delimiter_char": " ",
+  "target_lang": null,
+  "tokenizer_class": "Wav2Vec2CTCTokenizer",
+  "unk_token": "[UNK]",
+  "word_delimiter_token": "|"
+}

vocab.json

@@ -0,0 +1,78 @@
+{
+  "[PAD]": 75,
+  "[UNK]": 74,
+  "|": 13,
+  "ം": 2,
+  "ഃ": 39,
+  "അ": 48,
+  "ആ": 47,
+  "ഇ": 36,
+  "ഈ": 65,
+  "ഉ": 16,
+  "ഊ": 32,
+  "ഋ": 3,
+  "എ": 45,
+  "ഏ": 15,
+  "ഐ": 37,
+  "ഒ": 50,
+  "ഓ": 57,
+  "ഔ": 72,
+  "ക": 12,
+  "ഖ": 52,
+  "ഗ": 55,
+  "ഘ": 26,
+  "ങ": 68,
+  "ച": 41,
+  "ഛ": 30,
+  "ജ": 70,
+  "ഝ": 31,
+  "ഞ": 60,
+  "ട": 8,
+  "ഠ": 7,
+  "ഡ": 44,
+  "ഢ": 6,
+  "ണ": 19,
+  "ത": 69,
+  "ഥ": 43,
+  "ദ": 28,
+  "ധ": 5,
+  "ന": 33,
+  "പ": 9,
+  "ഫ": 53,
+  "ബ": 29,
+  "ഭ": 62,
+  "മ": 49,
+  "യ": 10,
+  "ര": 14,
+  "റ": 61,
+  "ല": 24,
+  "ള": 56,
+  "ഴ": 73,
+  "വ": 59,
+  "ശ": 51,
+  "ഷ": 46,
+  "സ": 1,
+  "ഹ": 4,
+  "ാ": 40,
+  "ി": 20,
+  "ീ": 11,
+  "ു": 54,
+  "ൂ": 58,
+  "ൃ": 17,
+  "െ": 67,
+  "േ": 42,
+  "ൈ": 38,
+  "ൊ": 23,
+  "ോ": 34,
+  "ൌ": 27,
+  "്": 66,
+  "ൗ": 25,
+  "൱": 35,
+  "ൺ": 18,
+  "ൻ": 63,
+  "ർ": 0,
+  "ൽ": 22,
+  "ൾ": 64,
+  "‌": 21,
+  "‍": 71
+}