added egyptian arabic model

README.md

@@ -0,0 +1,127 @@
+---
+language: arz
+datasets:
+- https://arabicspeech.org/
+metrics:
+- wer
+tags:
+- audio
+- automatic-speech-recognition
+- speech
+- xlsr-fine-tuning-week
+license: apache-2.0
+model-index:
+- name: `XLSR Wav2Vec2 Egyptian Arabic by Othmane Rifki`
+  results:
+  - task: 
+      name: Speech Recognition
+      type: automatic-speech-recognition
+    dataset:
+      name: arabicspeech.org MGB-3
+      type: arabicspeech.org MGB-3
+      args: ar
+    metrics:
+       - name: Test WER
+         type: wer
+         value: 55.2
+---
+
+# Wav2Vec2-Large-XLSR-53-Egyptian-Arabic
+
+Fine-tuned [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) in Egyptian using the [arabicspeech.org MGB-3](https://arabicspeech.org/mgb3-asr/)
+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", "ar", split="test[:2%]")
+
+processor = Wav2Vec2Processor.from_pretrained("othrif/wav2vec2-large-xlsr-egyptian")
+model = Wav2Vec2ForCTC.from_pretrained("othrif/wav2vec2-large-xlsr-egyptian")
+
+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"][:2])
+```
+
+
+## Evaluation
+
+The model can be evaluated as follows on the Arabic 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", "ar", split="test") 
+wer = load_metric("wer")
+
+processor = Wav2Vec2Processor.from_pretrained("othrif/wav2vec2-large-xlsr-egyptian") 
+model = Wav2Vec2ForCTC.from_pretrained("othrif/wav2vec2-large-xlsr-egyptian")
+model.to("cuda")
+
+chars_to_ignore_regex = '[\؛\—\_get\«\»\ـ\ـ\,\?\.\!\-\;\:\"\“\%\‘\”\�\#\،\☭,\؟]'
+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):
+	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 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**: 55.2
+
+
+## Training
+
+The Common Voice `train`, `validation` datasets were used for training.
+
+The script used for training can be found [here](https://github.com/othrif/xlsr-wav2vec2) 

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": 43,
+  "transformers_version": "4.4.0",
+  "vocab_size": 44
+}

optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0afe7313e3a58ee05f77356220557a12a7999c69667fff22cb67ac4ec07f91cc
+size 2490437895

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 sha256:455418f744560b3cf4356a8df6b0f13e984c0d73c886d371a7d1d61149befed7
+size 1262114199

scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0818666afb82b7b432824bd91f7b1cbfd826c0bbd672aea23cda693e38746b70
+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": 18.46153846153846,
+  "global_step": 1200,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 3.08,
+      "learning_rate": 0.00011999999999999999,
+      "loss": 5.8244,
+      "step": 200
+    },
+    {
+      "epoch": 3.08,
+      "eval_loss": 3.064483404159546,
+      "eval_runtime": 32.6688,
+      "eval_samples_per_second": 7.071,
+      "eval_wer": 0.9981722641078364,
+      "step": 200
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 3.0089,
+      "step": 400
+    },
+    {
+      "epoch": 6.15,
+      "eval_loss": 2.9369325637817383,
+      "eval_runtime": 32.3179,
+      "eval_samples_per_second": 7.148,
+      "eval_wer": 0.9981722641078364,
+      "step": 400
+    },
+    {
+      "epoch": 9.23,
+      "learning_rate": 0.0002625,
+      "loss": 2.2569,
+      "step": 600
+    },
+    {
+      "epoch": 9.23,
+      "eval_loss": 1.1511069536209106,
+      "eval_runtime": 32.4921,
+      "eval_samples_per_second": 7.109,
+      "eval_wer": 0.7900388393877085,
+      "step": 600
+    },
+    {
+      "epoch": 12.31,
+      "learning_rate": 0.00018749999999999998,
+      "loss": 0.9673,
+      "step": 800
+    },
+    {
+      "epoch": 12.31,
+      "eval_loss": 0.8750982284545898,
+      "eval_runtime": 32.4693,
+      "eval_samples_per_second": 7.114,
+      "eval_wer": 0.6248572081334247,
+      "step": 800
+    },
+    {
+      "epoch": 15.38,
+      "learning_rate": 0.0001125,
+      "loss": 0.6495,
+      "step": 1000
+    },
+    {
+      "epoch": 15.38,
+      "eval_loss": 0.8149241805076599,
+      "eval_runtime": 32.8858,
+      "eval_samples_per_second": 7.024,
+      "eval_wer": 0.5716244002741604,
+      "step": 1000
+    },
+    {
+      "epoch": 18.46,
+      "learning_rate": 3.75e-05,
+      "loss": 0.5093,
+      "step": 1200
+    },
+    {
+      "epoch": 18.46,
+      "eval_loss": 0.8335053324699402,
+      "eval_runtime": 32.889,
+      "eval_samples_per_second": 7.024,
+      "eval_wer": 0.5522047064199224,
+      "step": 1200
+    }
+  ],
+  "max_steps": 1300,
+  "num_train_epochs": 20,
+  "total_flos": 7.999569757266722e+18,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4a949b13ad5afbc0f1621dd3c424c1db200517da21931d0b7976930beff9c6f6
+size 2287

vocab.json

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