Upload 20 files

.gitattributes

@@ -32,3 +32,5 @@ saved_model/**/* 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
+vakyansh-wav2vec2-hindi-him-4200/language_model/kenLM.arpa filter=lfs diff=lfs merge=lfs -text
+vakyansh-wav2vec2-hindi-him-4200/language_model/unigrams.txt filter=lfs diff=lfs merge=lfs -text

alphabet.json

@@ -0,0 +1 @@
+{"labels": ["<s>", "", "</s>", "\u2047", " ", "\u0901", "\u0902", "\u0903", "\u0905", "\u0906", "\u0907", "\u0908", "\u0909", "\u090a", "\u090b", "\u090f", "\u0910", "\u0911", "\u0913", "\u0914", "\u0915", "\u0916", "\u0917", "\u0918", "\u0919", "\u091a", "\u091b", "\u091c", "\u091d", "\u091e", "\u091f", "\u0920", "\u0921", "\u0922", "\u0923", "\u0924", "\u0925", "\u0926", "\u0927", "\u0928", "\u092a", "\u092b", "\u092c", "\u092d", "\u092e", "\u092f", "\u0930", "\u0932", "\u0935", "\u0936", "\u0937", "\u0938", "\u0939", "\u093c", "\u093e", "\u093f", "\u0940", "\u0941", "\u0942", "\u0943", "\u0945", "\u0947", "\u0948", "\u0949", "\u094b", "\u094c", "\u094d"], "is_bpe": false}

language_model/Hin.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:66bd3713cbc78b75aa054e15bafc80b435879cfe81482868a5b1d889d0e0e0b4
+size 101236443

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,
+  "processor_class": "Wav2Vec2ProcessorWithLM",
+  "return_attention_mask": false,
+  "sampling_rate": 16000
+}

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 +1,14 @@
+{
+  "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": " ",
+  "tokenizer_class": "Wav2Vec2CTCTokenizer",
+  "trust_remote_code": false,
+  "unk_token": "<unk>",
+  "word_delimiter_token": "|"
+}

vakyansh-wav2vec2-hindi-him-4200/.gitattributes

@@ -0,0 +1,19 @@
+*.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
+*tfevents* filter=lfs diff=lfs merge=lfs -text
+language_model/kenLM.arpa filter=lfs diff=lfs merge=lfs -text
+language_model/unigrams.txt filter=lfs diff=lfs merge=lfs -text

vakyansh-wav2vec2-hindi-him-4200/README.md

@@ -0,0 +1,138 @@
+---
+language: hi
+#datasets:
+#- Interspeech 2021
+metrics:
+- wer
+tags:
+- audio
+- automatic-speech-recognition
+- speech
+license: mit
+model-index:
+- name: Wav2Vec2 Vakyansh Hindi Model by Harveen Chadha
+  results:
+  - task:
+      name: Speech Recognition
+      type: automatic-speech-recognition
+    dataset:
+      name: Common Voice hi
+      type: common_voice
+      args: hi
+    metrics:
+    - name: Test WER
+      type: wer
+      value: 33.17
+---
+
+## Spaces Demo
+Check the spaces demo [here](https://huggingface.co/spaces/Harveenchadha/wav2vec2-vakyansh-hindi/tree/main)
+
+## Pretrained Model
+
+Fine-tuned on Multilingual Pretrained Model [CLSRIL-23](https://arxiv.org/abs/2107.07402). The original fairseq checkpoint is present [here](https://github.com/Open-Speech-EkStep/vakyansh-models). When using this model, make sure that your speech input is sampled at 16kHz.
+
+**Note: The result from this model is without a language model so you may witness a higher WER in some cases.**
+
+## Dataset
+
+This model was trained on 4200 hours of Hindi Labelled Data. The labelled data is not present in public domain as of now.
+
+## Training Script
+
+Models were trained using experimental platform setup by Vakyansh team at Ekstep. Here is the [training repository](https://github.com/Open-Speech-EkStep/vakyansh-wav2vec2-experimentation).
+
+In case you want to explore training logs on wandb they are [here](https://wandb.ai/harveenchadha/hindi_finetuning_multilingual?workspace=user-harveenchadha).
+
+
+## [Colab Demo](https://colab.research.google.com/github/harveenchadha/bol/blob/main/demos/hf/hindi/hf_hindi_him_4200_demo.ipynb)
+
+## Usage
+
+The model can be used directly (without a language model) as follows:
+
+```python
+import soundfile as sf
+import torch
+from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
+import argparse
+
+def parse_transcription(wav_file):
+    # load pretrained model
+    processor = Wav2Vec2Processor.from_pretrained("Harveenchadha/vakyansh-wav2vec2-hindi-him-4200")
+    model = Wav2Vec2ForCTC.from_pretrained("Harveenchadha/vakyansh-wav2vec2-hindi-him-4200")
+
+    # load audio
+    audio_input, sample_rate = sf.read(wav_file)
+
+    # pad input values and return pt tensor
+    input_values = processor(audio_input, sampling_rate=sample_rate, return_tensors="pt").input_values
+
+    # INFERENCE
+    # retrieve logits & take argmax
+    logits = model(input_values).logits
+    predicted_ids = torch.argmax(logits, dim=-1)
+
+    # transcribe
+    transcription = processor.decode(predicted_ids[0], skip_special_tokens=True)
+    print(transcription)
+
+```
+
+
+## Evaluation
+The model can be evaluated as follows on the hindi test data of Common Voice. 
+
+```python
+
+import torch
+import torchaudio
+from datasets import load_dataset, load_metric
+from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
+import re
+
+test_dataset = load_dataset("common_voice", "hi", split="test")
+wer = load_metric("wer")
+
+processor = Wav2Vec2Processor.from_pretrained("Harveenchadha/vakyansh-wav2vec2-hindi-him-4200")
+model = Wav2Vec2ForCTC.from_pretrained("Harveenchadha/vakyansh-wav2vec2-hindi-him-4200")
+model.to("cuda")
+
+resampler = torchaudio.transforms.Resample(48_000, 16_000)
+
+chars_to_ignore_regex = '[\,\?\.\!\-\;\:\"\“]'
+
+# Preprocessing the datasets.
+# We need to read the aduio files as arrays
+def speech_file_to_array_fn(batch):
+  batch["sentence"] = re.sub(chars_to_ignore_regex, '', batch["sentence"]).lower()
+  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)
+
+# Preprocessing the datasets.
+# We need to read the aduio 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")).logits
+
+      pred_ids = torch.argmax(logits, dim=-1)
+      batch["pred_strings"] = processor.batch_decode(pred_ids, skip_special_tokens=True)
+      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**: 33.17 %
+
+[**Colab Evaluation**](https://colab.research.google.com/github/harveenchadha/bol/blob/main/demos/hf/hindi/hf_vakyansh_hindi_him_4200_evaluation_common_voice.ipynb) 
+
+## Credits
+Thanks to Ekstep Foundation for making this possible. The vakyansh team will be open sourcing speech models in all the Indic Languages.

vakyansh-wav2vec2-hindi-him-4200/alphabet.json

@@ -0,0 +1 @@
+{"labels": ["<s>", "", "</s>", "\u2047", " ", "\u0901", "\u0902", "\u0903", "\u0905", "\u0906", "\u0907", "\u0908", "\u0909", "\u090a", "\u090b", "\u090f", "\u0910", "\u0911", "\u0913", "\u0914", "\u0915", "\u0916", "\u0917", "\u0918", "\u0919", "\u091a", "\u091b", "\u091c", "\u091d", "\u091e", "\u091f", "\u0920", "\u0921", "\u0922", "\u0923", "\u0924", "\u0925", "\u0926", "\u0927", "\u0928", "\u092a", "\u092b", "\u092c", "\u092d", "\u092e", "\u092f", "\u0930", "\u0932", "\u0935", "\u0936", "\u0937", "\u0938", "\u0939", "\u093c", "\u093e", "\u093f", "\u0940", "\u0941", "\u0942", "\u0943", "\u0945", "\u0947", "\u0948", "\u0949", "\u094b", "\u094c", "\u094d"], "is_bpe": false}

vakyansh-wav2vec2-hindi-him-4200/config.json

@@ -0,0 +1,81 @@
+{
+  "activation_dropout": 0.1,
+  "apply_spec_augment": true,
+  "architectures": [
+    "Wav2Vec2ForCTC"
+  ],
+  "attention_dropout": 0.1,
+  "bos_token": "<s>",
+  "bos_token_id": 0,
+  "codevector_dim": 256,
+  "contrastive_logits_temperature": 0.1,
+  "conv_bias": false,
+  "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": "sum",
+  "ctc_zero_infinity": false,
+  "diversity_loss_weight": 0.1,
+  "do_lower_case": false,
+  "do_stable_layer_norm": false,
+  "eos_token": "</s>",
+  "eos_token_id": 2,
+  "feat_extract_activation": "gelu",
+  "feat_extract_norm": "group",
+  "feat_proj_dropout": 0.1,
+  "feat_quantizer_dropout": 0.0,
+  "final_dropout": 0.1,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-05,
+  "layerdrop": 0.1,
+  "mask_feature_length": 10,
+  "mask_feature_prob": 0.0,
+  "mask_time_length": 10,
+  "mask_time_prob": 0.05,
+  "model_type": "wav2vec2",
+  "num_attention_heads": 12,
+  "num_codevector_groups": 2,
+  "num_codevectors_per_group": 320,
+  "num_conv_pos_embedding_groups": 16,
+  "num_conv_pos_embeddings": 128,
+  "num_feat_extract_layers": 7,
+  "num_hidden_layers": 12,
+  "num_negatives": 100,
+  "pad_token": "[PAD]",
+  "pad_token_id": 1,
+  "proj_codevector_dim": 256,
+  "torch_dtype": "float32",
+  "transformers_version": "4.9.1",
+  "unk_token": "[UNK]",
+  "vocab_size": 67,
+  "word_delimiter_token": "|"
+}

vakyansh-wav2vec2-hindi-him-4200/language_model/attrs.json

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

vakyansh-wav2vec2-hindi-him-4200/language_model/kenLM.arpa

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec083e80a037ea5e4239974b2daae368ade0f89e9477f075cc0e15e0a77eb87c
+size 2008446585

vakyansh-wav2vec2-hindi-him-4200/language_model/unigrams.txt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7d5d4d03ad1c72ebc85c930a5fa7d5e0730a6d352a633f0fd0e5d93760053338
+size 11053241

vakyansh-wav2vec2-hindi-him-4200/preprocessor_config.json

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

vakyansh-wav2vec2-hindi-him-4200/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9bb46884e079e83cdc30e45b931f35f45bda2f4b74bf82fed1d2630ca25603a0
+size 377774679

vakyansh-wav2vec2-hindi-him-4200/special_tokens_map.json

@@ -0,0 +1,6 @@
+{
+  "bos_token": "<s>",
+  "eos_token": "</s>",
+  "pad_token": "<pad>",
+  "unk_token": "<unk>"
+}

vakyansh-wav2vec2-hindi-him-4200/tokenizer_config.json

@@ -0,0 +1,14 @@
+{
+  "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": " ",
+  "tokenizer_class": "Wav2Vec2CTCTokenizer",
+  "trust_remote_code": false,
+  "unk_token": "<unk>",
+  "word_delimiter_token": "|"
+}

vakyansh-wav2vec2-hindi-him-4200/vocab.json

@@ -0,0 +1,69 @@
+{
+  "</s>": 2,
+  "<pad>": 1,
+  "<s>": 0,
+  "<unk>": 3,
+  "|": 4,
+  "ँ": 5,
+  "ं": 6,
+  "ः": 7,
+  "अ": 8,
+  "आ": 9,
+  "इ": 10,
+  "ई": 11,
+  "उ": 12,
+  "ऊ": 13,
+  "ऋ": 14,
+  "ए": 15,
+  "ऐ": 16,
+  "ऑ": 17,
+  "ओ": 18,
+  "औ": 19,
+  "क": 20,
+  "ख": 21,
+  "ग": 22,
+  "घ": 23,
+  "ङ": 24,
+  "च": 25,
+  "छ": 26,
+  "ज": 27,
+  "झ": 28,
+  "ञ": 29,
+  "ट": 30,
+  "ठ": 31,
+  "ड": 32,
+  "ढ": 33,
+  "ण": 34,
+  "त": 35,
+  "थ": 36,
+  "द": 37,
+  "ध": 38,
+  "न": 39,
+  "प": 40,
+  "फ": 41,
+  "ब": 42,
+  "भ": 43,
+  "म": 44,
+  "य": 45,
+  "र": 46,
+  "ल": 47,
+  "व": 48,
+  "श": 49,
+  "ष": 50,
+  "स": 51,
+  "ह": 52,
+  "़": 53,
+  "ा": 54,
+  "ि": 55,
+  "ी": 56,
+  "ु": 57,
+  "ू": 58,
+  "ृ": 59,
+  "ॅ": 60,
+  "े": 61,
+  "ै": 62,
+  "ॉ": 63,
+  "ो": 64,
+  "ौ": 65,
+  "्": 66
+}

vocab.json

@@ -0,0 +1,69 @@
+{
+  "</s>": 2,
+  "<pad>": 1,
+  "<s>": 0,
+  "<unk>": 3,
+  "|": 4,
+  "ँ": 5,
+  "ं": 6,
+  "ः": 7,
+  "अ": 8,
+  "आ": 9,
+  "इ": 10,
+  "ई": 11,
+  "उ": 12,
+  "ऊ": 13,
+  "ऋ": 14,
+  "ए": 15,
+  "ऐ": 16,
+  "ऑ": 17,
+  "ओ": 18,
+  "औ": 19,
+  "क": 20,
+  "ख": 21,
+  "ग": 22,
+  "घ": 23,
+  "ङ": 24,
+  "च": 25,
+  "छ": 26,
+  "ज": 27,
+  "झ": 28,
+  "ञ": 29,
+  "ट": 30,
+  "ठ": 31,
+  "ड": 32,
+  "ढ": 33,
+  "ण": 34,
+  "त": 35,
+  "थ": 36,
+  "द": 37,
+  "ध": 38,
+  "न": 39,
+  "प": 40,
+  "फ": 41,
+  "ब": 42,
+  "भ": 43,
+  "म": 44,
+  "य": 45,
+  "र": 46,
+  "ल": 47,
+  "व": 48,
+  "श": 49,
+  "ष": 50,
+  "स": 51,
+  "ह": 52,
+  "़": 53,
+  "ा": 54,
+  "ि": 55,
+  "ी": 56,
+  "ु": 57,
+  "ू": 58,
+  "ृ": 59,
+  "ॅ": 60,
+  "े": 61,
+  "ै": 62,
+  "ॉ": 63,
+  "ो": 64,
+  "ौ": 65,
+  "्": 66
+}