Upload README.md with huggingface_hub

README.md

@@ -14,33 +14,52 @@ pipeline_tag: automatic-speech-recognition
 
 # whisper-base-it
 
-Fine-tuned [openai/whisper-base](https://huggingface.co/openai/whisper-base) for Italian automatic speech recognition (ASR).
+Fine-tuned [openai/whisper-base](https://huggingface.co/openai/whisper-base) (74M params) for Italian automatic speech recognition (ASR).
 
 **Author:** Ettore Di Giacinto
 
 Brought to you by the [LocalAI](https://github.com/mudler/LocalAI) team. This model can be used directly with [LocalAI](https://localai.io).
 
+## Results
+
+Evaluated on Common Voice 25.0 Italian test set (15,184 samples):
+
+| Step | WER |
+|------|-----|
+| 1000 | 26.5% |
+| 2000 | 24.0% |
+| 3000 | 22.4% |
+| 5000 | 20.6% |
+| 7000 | 19.9% |
+| 10000 | **19.2%** |
+
 ## Training Details
 
-- **Base model:** openai/whisper-base
+- **Base model:** openai/whisper-base (74M parameters)
 - **Dataset:** Common Voice 25.0 Italian (173k train, 15k dev, 15k test)
-- **Steps:** 10,000
-- **Precision:** bf16
+- **Steps:** 10,000 (batch size 32, ~1.8 epochs)
+- **Learning rate:** 1e-5 with 500 warmup steps
+- **Precision:** bf16 on NVIDIA GB10
 
 ## Usage
 
 ### Transformers
 
-\`\`\`python
+```python
 from transformers import pipeline
+
 pipe = pipeline("automatic-speech-recognition", model="LocalAI-io/whisper-base-it")
 result = pipe("audio.mp3", generate_kwargs={"language": "it", "task": "transcribe"})
 print(result["text"])
-\`\`\`
+```
+
+### CTranslate2 / faster-whisper
+
+For optimized CPU inference, use the INT8 quantized version: [LocalAI-io/whisper-base-it-ct2-int8](https://huggingface.co/LocalAI-io/whisper-base-it-ct2-int8) (79MB).
 
-### CTranslate2 / faster-whisper (INT8, optimized for CPU)
+### LocalAI
 
-See: [LocalAI-io/whisper-base-it-ct2-int8](https://huggingface.co/LocalAI-io/whisper-base-it-ct2-int8)
+This model is compatible with [LocalAI](https://github.com/mudler/LocalAI) for local, self-hosted AI inference.
 
 ## Links