Upload 6 files

README.md

@@ -0,0 +1,122 @@
+---
+language: or
+datasets:
+- OpenSLR
+metrics:
+- wer
+tags:
+- audio
+- automatic-speech-recognition
+- speech
+- xlsr-fine-tuning-week
+license: apache-2.0
+model-index:
+- name: XLSR Wav2Vec2 Odia by Shyam Sunder Kumar
+  results:
+  - task: 
+      name: Speech Recognition
+      type: automatic-speech-recognition
+    dataset:
+      name: OpenSLR
+      type: OpenSLR
+      args: or
+    metrics:
+       - name: Test WER
+         type: wer
+         value: 68.75
+---
+
+# Wav2Vec2-Large-XLSR-53-Odia
+
+Fine-tuned [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) odia using the [Multilingual and code-switching ASR challenges for low resource Indian languages](https://navana-tech.github.io/IS21SS-indicASRchallenge/data.html).
+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_dataset("common_voice", "or", split="test[:2%]")
+processor = Wav2Vec2Processor.from_pretrained("theainerd/wav2vec2-large-xlsr-53-odia")
+model = Wav2Vec2ForCTC.from_pretrained("theainerd/wav2vec2-large-xlsr-53-odia")
+resampler = torchaudio.transforms.Resample(48_000, 16_000)
+
+# Preprocessing the datasets.
+# We need to read the aduio 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"][:2])
+```
+
+
+## Evaluation
+
+The model can be evaluated as follows on the Odia 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", "or", split="test")
+wer = load_metric("wer")
+
+processor = Wav2Vec2Processor.from_pretrained("theainerd/wav2vec2-large-xlsr-53-odia")
+model = Wav2Vec2ForCTC.from_pretrained("theainerd/wav2vec2-large-xlsr-53-odia")
+model.to("cuda")
+
+resampler = torchaudio.transforms.Resample(48_000, 16_000)
+
+# 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"), 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**: 68.75 %
+
+
+## Training
+
+The script used for training can be found [Odia ASR Fine Tuning Wav2Vec2](https://colab.research.google.com/drive/1aHpFRTxaBeNblRHAtYOy0hBeXbbMWtot?usp=sharing)

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": 66,
+  "transformers_version": "4.5.0.dev0",
+  "vocab_size": 67
+}

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
+}

special_tokens_map.json

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

tokenizer_config.json

@@ -0,0 +1 @@
+{"unk_token": "[UNK]", "bos_token": "<s>", "eos_token": "</s>", "pad_token": "[PAD]", "do_lower_case": false, "word_delimiter_token": "|", "special_tokens_map_file": "/content/drive/MyDrive/ASR_models/wav2vec2-large-xlsr-odia/special_tokens_map.json", "tokenizer_file": null, "name_or_path": "/content/drive/MyDrive/ASR_models/wav2vec2-large-xlsr-odia"}

vocab.json

@@ -0,0 +1 @@
+{"ପ": 0, "ଣ": 1, "ି": 2, "ଟ": 3, "ଥ": 4, "ୂ": 5, "ଡ": 6, "ଋ": 7, "ଏ": 8, "ଔ": 9, "ଘ": 10, "ଠ": 11, "ଈ": 12, "ୋ": 13, "ଝ": 14, "ଆ": 15, "ତ": 16, "ର": 17, "ୈ": 18, "ହ": 19, "ବ": 20, "କ": 21, "ୃ": 22, "ଯ": 23, "ଞ": 24, "ଳ": 25, "ଉ": 26, "ଇ": 27, "ଅ": 28, "ଷ": 29, "ଜ": 30, ".": 31, "ଛ": 32, "ଭ": 33, "ା": 34, "ୌ": 35, "ଦ": 36, "ଂ": 37, "ୀ": 38, "ଶ": 39, "ଧ": 40, "ଗ": 41, "ଃ": 42, "ୱ": 43, "ଙ": 44, "଼": 45, "ୁ": 46, "ଖ": 47, "ମ": 48, "ଁ": 49, "ଚ": 50, "ୟ": 51, "ସ": 52, "ଵ": 53, "ଊ": 54, "ଫ": 55, "େ": 56, "ଓ": 57, "ଢ": 58, "ଢ଼": 59, "ନ": 60, "ଲ": 62, "ଡ଼": 63, "୍": 64, "|": 61, "[UNK]": 65, "[PAD]": 66}