Removed demo file from README and removed INT8 references from README for now

INT8 feature was not working fully, need to investigate more before applying again to this repository.

README.md

@@ -1,5 +1,5 @@
 ---
-title: NVIDIA Parakeet TDT 0.6B V2 INT8 Real Time Mic Transcription
+title: NVIDIA Parakeet TDT 0.6B V2 Real Time Mic Transcription
 emoji: 📊
 colorFrom: purple
 colorTo: blue
@@ -17,7 +17,6 @@ tags:
   - speech-recognition
   - asr
   - real-time
-  - int8
   - cpu
   - nvidia
   - parakeet
@@ -30,10 +29,10 @@ tags:
   - huggingface
 ---
 
-# 🦜 NVIDIA Parakeet-TDT-0.6B-v2 (INT8) — CPU-Only Streaming ASR
+# 🦜 NVIDIA Parakeet-TDT-0.6B-v2 — CPU-Only Streaming ASR
 
 **Real-time English speech-to-text in your browser — no GPU required.**  
-This Space runs the 600 M-parameter [`nvidia/parakeet-tdt-0.6b-v2`](https://huggingface.co/nvidia/parakeet-tdt-0.6b-v2) model with dynamic **INT8 quantization** so it fits comfortably on the **CPU Basic (2 vCPU)** tier.
+This Space runs the 600 M-parameter [`nvidia/parakeet-tdt-0.6b-v2`](https://huggingface.co/nvidia/parakeet-tdt-0.6b-v2) model that fits comfortably on the **CPU Basic (2 vCPU)** tier.
 
 ## 🚀 Quick Start
 1. Click **“Record”**
@@ -42,10 +41,6 @@ This Space runs the 600 M-parameter [`nvidia/parakeet-tdt-0.6b-v2`](https://hugg
 
 > **Stalled UI?** Refresh the browser tab — this fully restarts the Space and clears any stuck threads.
 
-<video src="https://huggingface.co/spaces/WJ88/NVIDIA-Parakeet-TDT-0.6B-v2-INT8-Real-Time-Mic-Transcription/resolve/main/demo0__5-24-2025.mp4" controls style="max-width: 100%; height: auto;">
-  Your browser does not support the video tag.
-</video>
-
 ## 🔧 Build on This
 - **Duplicate** the Space (button at the top-right) to kick-start your own ASR ideas.  
 - Swap in another NeMo/HF model — the quantization + streaming scaffold is ready.  
@@ -54,9 +49,8 @@ This Space runs the 600 M-parameter [`nvidia/parakeet-tdt-0.6b-v2`](https://hugg
 ## ⚙️ Under the Hood
 | Technique | Why it matters |
 |-----------|----------------|
-| **Dynamic INT8 quantization** (`torch.quantization.quantize_dynamic`) | ~4× smaller, faster CPU inference with minimal accuracy loss |
 | **`OMP_NUM_THREADS=2` & `torch.set_num_threads(2)`** | Matches the 2 vCPUs for optimal throughput |
-| **FBGEMM backend** | Fastest INT8 kernels on x86 |
+| **FBGEMM backend** | Fastest kernels on x86 |
 | **4-second streaming window** | Low latency & small memory footprint |
 | **Gradio `stream_every=0.5`** | Updates the transcript twice per second for real-time feel |
 
@@ -70,4 +64,4 @@ Feel free to browse `app.py` for the full implementation.
 
 If you redistribute transcripts or fine-tuned weights, please retain the CC-BY-4.0 attribution notice.
 
-⭐ **If this Space helps you, please give it a like and share your feedback!**
+⭐ **If this Space helps you, please give it a like and share your feedback!**