Upload README.md with huggingface_hub

README.md

@@ -1,76 +1,35 @@
 ---
-library_name: transformers
+language: ar
 license: apache-2.0
-base_model: tarteel-ai/whisper-base-ar-quran
 tags:
-- generated_from_trainer
+- automatic-speech-recognition
+- whisper
+- quran
+- warsh
+- arabic
+- tajweed
+- tahkik
+datasets:
+- benhadjermed/warsh_quran_aligned
+base_model: tarteel-ai/whisper-base-ar-quran
 metrics:
 - wer
-model-index:
-- name: tahkik-basic-warsh
-  results: []
+pipeline_tag: automatic-speech-recognition
 ---
 
-<!-- This model card has been generated automatically according to the information the Trainer had access to. You
-should probably proofread and complete it, then remove this comment. -->
-
-# tahkik-basic-warsh
-
-This model is a fine-tuned version of [tarteel-ai/whisper-base-ar-quran](https://huggingface.co/tarteel-ai/whisper-base-ar-quran) on an unknown dataset.
-It achieves the following results on the evaluation set:
-- Loss: 0.1363
-- Wer: 0.2579
-
-## Model description
-
-More information needed
-
-## Intended uses & limitations
-
-More information needed
-
-## Training and evaluation data
-
-More information needed
-
-## Training procedure
-
-### Training hyperparameters
-
-The following hyperparameters were used during training:
-- learning_rate: 1e-05
-- train_batch_size: 8
-- eval_batch_size: 8
-- seed: 42
-- gradient_accumulation_steps: 4
-- total_train_batch_size: 32
-- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
-- lr_scheduler_type: linear
-- lr_scheduler_warmup_steps: 200
-- training_steps: 3000
-- mixed_precision_training: Native AMP
-
-### Training results
-
-| Training Loss | Epoch  | Step | Validation Loss | Wer    |
-|:-------------:|:------:|:----:|:---------------:|:------:|
-| 0.9224        | 0.7128 | 250  | 0.2397          | 0.3444 |
-| 0.7034        | 1.4248 | 500  | 0.1826          | 0.2841 |
-| 0.5101        | 2.1368 | 750  | 0.1638          | 0.2719 |
-| 0.5467        | 2.8496 | 1000 | 0.1525          | 0.2636 |
-| 0.4900        | 3.5617 | 1250 | 0.1468          | 0.2568 |
-| 0.3796        | 4.2737 | 1500 | 0.1433          | 0.2614 |
-| 0.3936        | 4.9865 | 1750 | 0.1379          | 0.2593 |
-| 0.3418        | 5.6985 | 2000 | 0.1361          | 0.2604 |
-| 0.2647        | 6.4105 | 2250 | 0.1364          | 0.2605 |
-| 0.2935        | 7.1226 | 2500 | 0.1351          | 0.2576 |
-| 0.2669        | 7.8354 | 2750 | 0.1339          | 0.2565 |
-| 0.2836        | 8.5474 | 3000 | 0.1346          | 0.2601 |
+# Tahkik — Warsh Quran ASR
 
+Fine-tuned from [`tarteel-ai/whisper-base-ar-quran`](https://huggingface.co/tarteel-ai/whisper-base-ar-quran)
+on [`benhadjermed/warsh_quran_aligned`](https://huggingface.co/datasets/benhadjermed/warsh_quran_aligned).
 
-### Framework versions
+## Usage
+```python
+from transformers import pipeline
+asr = pipeline("automatic-speech-recognition", model="benhadjermed/tahkik-basic-warsh")
+print(asr("your_warsh_recitation.wav")["text"])
+```
 
-- Transformers 5.0.0
-- Pytorch 2.10.0+cu128
-- Datasets 4.0.0
-- Tokenizers 0.22.2
+## Training
+- Base model: `tarteel-ai/whisper-base-ar-quran`
+- Steps: 3000 | LR: 1e-05 | Effective batch: 32
+- Test WER: 0.2579