Upload Initial Model

README.md

@@ -0,0 +1,159 @@
+---
+language: ar
+datasets:
+- common_voice
+metrics:
+- wer
+tags:
+- audio
+- automatic-speech-recognition
+- speech
+- xlsr-fine-tuning-week
+license: apache-2.0
+model-index:
+- name: Sinai Voice Arabic Specch Recognition Model
+  results:
+  - task: 
+      name: Speech Recognition
+      type: automatic-speech-recognition
+    dataset:
+      name: Common Voice ar
+      type: common_voice
+      args: ar
+    metrics:
+       - name: Test WER
+         type: wer
+         value: 40.2
+---
+
+# Sinai Voice Arabic Speech Recognition Model
+# نموذج **صوت سيناء** للتعرف على الأصوات العربية الفصحى و تحويلها إلى نصوص
+Fine-tuned [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53)
+on Arabic using the [Common Voice](https://huggingface.co/datasets/common_voice)
+
+
+## Usage
+
+Please install:
+- [PyTorch](https://pytorch.org/)
+- `$ pip3 install jiwer lang_trans torchaudio datasets transformers`
+
+The model can be used directly (without a language model) as follows:
+```python
+import torch
+import torchaudio
+from datasets import load_dataset
+from lang_trans.arabic import buckwalter
+from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
+dataset = load_dataset("common_voice", "ar", split="test[:10]")
+resamplers = {  # all three sampling rates exist in test split
+    48000: torchaudio.transforms.Resample(48000, 16000),
+    44100: torchaudio.transforms.Resample(44100, 16000),
+    32000: torchaudio.transforms.Resample(32000, 16000),
+}
+def prepare_example(example):
+    speech, sampling_rate = torchaudio.load(example["path"])
+    example["speech"] = resamplers[sampling_rate](speech).squeeze().numpy()
+    return example
+dataset = dataset.map(prepare_example)
+processor = Wav2Vec2Processor.from_pretrained("bakrianoo/sinai-voice-ar-stt")
+model = Wav2Vec2ForCTC.from_pretrained("bakrianoo/sinai-voice-ar-stt").eval()
+def predict(batch):
+    inputs = processor(batch["speech"], sampling_rate=16000, return_tensors="pt", padding=True)
+    with torch.no_grad():
+        predicted = torch.argmax(model(inputs.input_values).logits, dim=-1)
+    predicted[predicted == -100] = processor.tokenizer.pad_token_id  # see fine-tuning script
+    batch["predicted"] = processor.tokenizer.batch_decode(predicted)
+    return batch
+dataset = dataset.map(predict, batched=True, batch_size=1, remove_columns=["speech"])
+for reference, predicted in zip(dataset["sentence"], dataset["predicted"]):
+    print("reference:", reference)
+    print("predicted:", predicted)
+    print("--")
+```
+Here's the output:
+```
+reference: ألديك قلم ؟
+predicted: ألديك قلم
+--
+reference: ليست هناك مسافة على هذه الأرض أبعد من يوم أمس.
+predicted: ليست نارك مسافة على هذه الأرض أبعد من يوم أمس
+--
+reference: إنك تكبر المشكلة.
+predicted: إنك تكبر المشكلة
+--
+reference: يرغب أن يلتقي بك.
+predicted: يرغب أن يلتقي بك
+--
+reference: إنهم لا يعرفون لماذا حتى.
+predicted: إنهم لا يعرفون لماذا حتى
+--
+reference: سيسعدني مساعدتك أي وقت تحب.
+predicted: سيسعدن مساعثتك أي وقد تحب
+--
+reference: أَحَبُّ نظريّة علمية إليّ هي أن حلقات زحل مكونة بالكامل من الأمتعة المفقودة.
+predicted: أحب نظرية علمية إلي هي أن أحلقتز حلم كوينا بالكامل من الأمت عن المفقودة
+--
+reference: سأشتري له قلماً.
+predicted: سأشتري له قلما
+--
+reference: أين المشكلة ؟
+predicted: أين المشكل
+--
+reference: وَلِلَّهِ يَسْجُدُ مَا فِي السَّمَاوَاتِ وَمَا فِي الْأَرْضِ مِنْ دَابَّةٍ وَالْمَلَائِكَةُ وَهُمْ لَا يَسْتَكْبِرُونَ
+predicted: ولله يسجد ما في السماوات وما في الأرض من دابة والملائكة وهم لا يستكبرون
+```
+## Evaluation
+
+CLONED from [elgeish/wav2vec2-large-xlsr-53-arabic](https://huggingface.co/elgeish/wav2vec2-large-xlsr-53-arabic)
+
+The model can be evaluated as follows on the Arabic test data of Common Voice:
+```python
+import jiwer
+import torch
+import torchaudio
+from datasets import load_dataset
+from lang_trans.arabic import buckwalter
+from transformers import set_seed, Wav2Vec2ForCTC, Wav2Vec2Processor
+set_seed(42)
+test_split = load_dataset("common_voice", "ar", split="test")
+resamplers = {  # all three sampling rates exist in test split
+    48000: torchaudio.transforms.Resample(48000, 16000),
+    44100: torchaudio.transforms.Resample(44100, 16000),
+    32000: torchaudio.transforms.Resample(32000, 16000),
+}
+def prepare_example(example):
+    speech, sampling_rate = torchaudio.load(example["path"])
+    example["speech"] = resamplers[sampling_rate](speech).squeeze().numpy()
+    return example
+test_split = test_split.map(prepare_example)
+processor = Wav2Vec2Processor.from_pretrained("bakrianoo/sinai-voice-ar-stt")
+model = Wav2Vec2ForCTC.from_pretrained("bakrianoo/sinai-voice-ar-stt").to("cuda").eval()
+def predict(batch):
+    inputs = processor(batch["speech"], sampling_rate=16000, return_tensors="pt", padding=True)
+    with torch.no_grad():
+        predicted = torch.argmax(model(inputs.input_values.to("cuda")).logits, dim=-1)
+    predicted[predicted == -100] = processor.tokenizer.pad_token_id  # see fine-tuning script
+    batch["predicted"] = processor.batch_decode(predicted)
+    return batch
+test_split = test_split.map(predict, batched=True, batch_size=16, remove_columns=["speech"])
+transformation = jiwer.Compose([
+    # normalize some diacritics, remove punctuation, and replace Persian letters with Arabic ones
+    jiwer.SubstituteRegexes({
+        r'[auiFNKo\~_،؟»\?;:\-,\.؛«!"]': "", "\u06D6": "",
+        r"[\|\{]": "A", "p": "h", "ک": "k", "ی": "y"}),
+    # default transformation below
+    jiwer.RemoveMultipleSpaces(),
+    jiwer.Strip(),
+    jiwer.SentencesToListOfWords(),
+    jiwer.RemoveEmptyStrings(),
+])
+metrics = jiwer.compute_measures(
+    truth=[buckwalter.trans(s) for s in test_split["sentence"]],  # Buckwalter transliteration
+    hypothesis=test_split["predicted"],
+    truth_transform=transformation,
+    hypothesis_transform=transformation,
+)
+print(f"WER: {metrics['wer']:.2%}")
+```
+**Test Result**: 40.2%

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": 0.15,
+  "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": 0.15,
+  "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": 45,
+  "transformers_version": "4.4.0",
+  "vocab_size": 48
+}

optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ff523d662d9db49ccaadba49574509f044262f42534b15281436d469f3d2a65e
+size 2490464146

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:fe340e53abd873e326771a32ab2b1d21f4ae07a98401dd0ada85940dbacc6e92
+size 1262126414

scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7db558e13c832cb0475ee05ac56dc0ed99b68cb98e8687b128d7b06e57d53360
+size 623

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": "|"}

trainer_state.json

@@ -0,0 +1,100 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 21.238938053097346,
+  "global_step": 2400,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 3.54,
+      "learning_rate": 0.0001926605504587156,
+      "loss": NaN,
+      "step": 400
+    },
+    {
+      "epoch": 3.54,
+      "eval_loss": 2.914095878601074,
+      "eval_runtime": 779.3384,
+      "eval_samples_per_second": 9.78,
+      "eval_wer": 1.0,
+      "step": 400
+    },
+    {
+      "epoch": 7.08,
+      "learning_rate": 0.0001779816513761468,
+      "loss": NaN,
+      "step": 800
+    },
+    {
+      "epoch": 7.08,
+      "eval_loss": 0.5257614850997925,
+      "eval_runtime": 808.0741,
+      "eval_samples_per_second": 9.432,
+      "eval_wer": 0.4909435120753172,
+      "step": 800
+    },
+    {
+      "epoch": 10.62,
+      "learning_rate": 0.00016330275229357798,
+      "loss": NaN,
+      "step": 1200
+    },
+    {
+      "epoch": 10.62,
+      "eval_loss": 0.4604354500770569,
+      "eval_runtime": 816.9678,
+      "eval_samples_per_second": 9.33,
+      "eval_wer": 0.444765656979124,
+      "step": 1200
+    },
+    {
+      "epoch": 14.16,
+      "learning_rate": 0.00014862385321100919,
+      "loss": NaN,
+      "step": 1600
+    },
+    {
+      "epoch": 14.16,
+      "eval_loss": 0.4556906819343567,
+      "eval_runtime": 788.8971,
+      "eval_samples_per_second": 9.662,
+      "eval_wer": 0.4178008595988539,
+      "step": 1600
+    },
+    {
+      "epoch": 17.7,
+      "learning_rate": 0.00013394495412844036,
+      "loss": NaN,
+      "step": 2000
+    },
+    {
+      "epoch": 17.7,
+      "eval_loss": 0.44174715876579285,
+      "eval_runtime": 788.0029,
+      "eval_samples_per_second": 9.673,
+      "eval_wer": 0.4084885386819484,
+      "step": 2000
+    },
+    {
+      "epoch": 21.24,
+      "learning_rate": 0.00011926605504587157,
+      "loss": NaN,
+      "step": 2400
+    },
+    {
+      "epoch": 21.24,
+      "eval_loss": 0.43745675683021545,
+      "eval_runtime": 797.2999,
+      "eval_samples_per_second": 9.56,
+      "eval_wer": 0.4028602128530495,
+      "step": 2400
+    }
+  ],
+  "max_steps": 5650,
+  "num_train_epochs": 50,
+  "total_flos": 8.638551360908018e+19,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b674eec74b835dba17b22c39ba4f76f198e2fd26941f1588e8a63aa2c526e6f9
+size 2287

vocab.json

@@ -0,0 +1 @@
+{"ج": 0, "ح": 1, "ﺃ": 2, "ت": 3, "ط": 4, "خ": 5, "چ": 6, "س": 7, "ب": 8, "غ": 10, "ث": 11, "ض": 12, "ا": 13, "ذ": 14, "ھ": 15, "ز": 16, "ى": 17, "ﻻ": 18, "ظ": 19, "ق": 20, "ص": 21, "م": 22, "ف": 23, "د": 24, "ش": 25, "و": 26, "ه": 27, "ی": 28, "ء": 29, "ر": 30, "آ": 31, "ع": 32, "ي": 33, "ل": 34, "ؤ": 35, "ڨ": 36, "ک": 37, "إ": 38, "أ": 39, "ك": 40, "ة": 41, "ئ": 42, "ن": 43, "|": 9, "[UNK]": 44, "[PAD]": 45}