Card: align framing to the technical report (open-weight foundation, adaptation-not-proprietary), add report link

README.md

@@ -27,7 +27,7 @@ tags:
 
 # Audar-ASR-V1-Turbo · GGUF
 
-### Audar's proprietary Arabic speech-recognition model — leaderboard-grade, dialect-aware.
+### Audar's Arabic-first speech-recognition model — leaderboard-grade, dialect-aware.
 
 **From Arabic to the world.**
 
@@ -40,7 +40,7 @@ tags:
 ![Emirati](https://img.shields.io/badge/Emirati-19.4%25%20WER%20%2F%207.3%25%20CER-brightgreen)
 [![GitHub](https://img.shields.io/badge/GitHub-Audar--ASR--V1-181717?logo=github)](https://github.com/AudarAI/Audar-ASR-V1)
 
-<p><a href="#-what-it-is"><b>🧭 Overview</b></a> · <a href="#-benchmarks"><b>📊 Benchmarks</b></a> · <a href="#-gguf-inference-llamacpp"><b>💻 GGUF Deploy</b></a> · <a href="#-real-time-streaming"><b>🎙️ Streaming</b></a> · <a href="#-vllm-inference-gpu-serving"><b>⚡ vLLM</b></a> · <a href="https://github.com/AudarAI/Audar-ASR-V1"><b>🐙 GitHub</b></a> · <a href="https://www.audarai.com"><b>☁️ Audar API</b></a> · <a href="https://www.audarai.com/license/audarai-community-license-v1.0/"><b>📜 License</b></a></p>
+<p><a href="#-what-it-is"><b>🧭 Overview</b></a> · <a href="#-benchmarks"><b>📊 Benchmarks</b></a> · <a href="#-gguf-inference-llamacpp"><b>💻 GGUF Deploy</b></a> · <a href="#-real-time-streaming"><b>🎙️ Streaming</b></a> · <a href="#-vllm-inference-gpu-serving"><b>⚡ vLLM</b></a> · <a href="https://github.com/AudarAI/Audar-ASR-V1/blob/main/report/Audar-ASR-V1-Technical-Report.pdf"><b>📄 Tech Report</b></a> · <a href="https://github.com/AudarAI/Audar-ASR-V1"><b>🐙 GitHub</b></a> · <a href="https://www.audarai.com"><b>☁️ Audar API</b></a> · <a href="https://www.audarai.com/license/audarai-community-license-v1.0/"><b>📜 License</b></a></p>
 
 </div>
 
@@ -48,12 +48,12 @@ tags:
 
 ## 🧭 What it is
 
-**Audar-ASR-V1-Turbo** is **Audar's proprietary Arabic speech-recognition model** — the accuracy tier of
+**Audar-ASR-V1-Turbo** is an **Arabic-first generative speech-recognition model** — the accuracy tier of
 the Audar-ASR family. It recasts transcription as **audio-conditioned next-token prediction** over a
 unified text vocabulary (a language-model decoder rather than a CTC or transducer objective), and is
-developed **in-house** through a proprietary Arabic training program:
+built on a permissively-licensed open-weight audio-LLM foundation and adapted in-house — the contribution is the adaptation (the data curriculum and the alignment rubric), not the foundation:
 
-- 🧱 **Large-scale dialectal pretraining** — 300,000+ hours of Arabic audio spanning MSA, Gulf,
+- 🧱 **Large-scale bilingual pretraining** — 300,000+ hours of labeled audio, primarily Arabic and English, spanning MSA, Gulf,
   Egyptian, Levantine and Maghrebi speech, code-switching, and diverse acoustic channels.
 - 🎯 **Dialect-targeted fine-tuning** — hardness sampling and multi-task conditioning focused on proper
   nouns, code-switching, and dialect-faithful orthography.
@@ -66,17 +66,17 @@ The result is **state-of-the-art dialectal Arabic ASR** — the lowest average W
 evaluated system on the *Open Universal Arabic ASR Leaderboard*. It transcribes MSA and every major
 Arabic dialect, code-switched Arabic–English, and English, across **30 languages** in total.
 
-> Distributed in the widely-supported **Qwen3-ASR architecture format** for turnkey tooling
-> (llama.cpp / GGUF). The **model** — data, training curriculum, and alignment — is Audar's.
+> Built on a **permissively-licensed open-weight audio-LLM foundation**; the adaptation, data, and
+> alignment are Audar's. Full method and results: [Audar-ASR-V1 Technical Report](https://github.com/AudarAI/Audar-ASR-V1/blob/main/report/Audar-ASR-V1-Technical-Report.pdf).
 
 ## Model summary
 
 <table>
 <tbody>
-<tr><td width="200"><b>Model</b></td><td>Audar-ASR-V1-Turbo — proprietary Arabic ASR (accuracy tier)</td></tr>
+<tr><td width="200"><b>Model</b></td><td>Audar-ASR-V1-Turbo — Arabic-first generative ASR (accuracy tier)</td></tr>
 <tr><td><b>Task</b></td><td>Automatic speech recognition (audio → text)</td></tr>
 <tr><td><b>Approach</b></td><td>Generative ASR — audio encoder + language-model decoder (audio-conditioned next-token prediction)</td></tr>
-<tr><td><b>Training</b></td><td>300k+ hrs dialectal pretraining → dialect-targeted SFT → KTO preference alignment</td></tr>
+<tr><td><b>Training</b></td><td>built on an open-weight audio-LLM foundation; adapted via a 4-stage curriculum — 300k+ hrs bilingual pretraining → multi-task fine-tuning → dialect PEFT → KTO alignment</td></tr>
 <tr><td><b>Decoder parameters</b></td><td>2,031,739,904 (2.03B)</td></tr>
 <tr><td><b>Audio encoder parameters</b></td><td>317,477,504 (0.32B)</td></tr>
 <tr><td><b>Total parameters</b></td><td>2,349,217,408 (2.35B, bf16)</td></tr>
@@ -325,11 +325,11 @@ License. See
 
 ```bibtex
 @misc{audar-asr-turbo-2026,
-  title  = {Audar-ASR: Dialect-Aware Arabic Speech Recognition},
+  title  = {Audar-ASR-V1: A Multilingual, Arabic-First Generative Speech Recognition Foundation Model},
   author = {AudarAI},
   year   = {2026},
   note   = {Audar-ASR-V1-Turbo},
-  url    = {https://huggingface.co/audarai/Audar-ASR-V1-Turbo}
+  url    = {https://github.com/AudarAI/Audar-ASR-V1/blob/main/report/Audar-ASR-V1-Technical-Report.pdf}
 }
 ```
 

vllm-w4a16/configuration_audar_asr.py

@@ -23,10 +23,10 @@ class Qwen3ASRAudioEncoderConfig(PretrainedConfig):
     r"""
     This is the configuration class to store the configuration of a [`Qwen3ASRAudioEncoder`]. It is used to instantiate a
     Qwen3-ASR audio encoder according to the specified arguments, defining the model architecture. Instantiating a
-    configuration with the defaults will yield a similar configuration to that of the audio encoder of the Qwen2-Audio
+    configuration with the defaults will yield a similar configuration to that of the audio encoder of Audar-ASR-V1-Turbo
     architecture.
 
-    e.g. [Qwen/Qwen3-ASR-1.7B](https://huggingface.co/Qwen/Qwen3-ASR-1.7B)
+    e.g. Audar-ASR-V1-Turbo
 
     Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
     documentation from [`PretrainedConfig`] for more information.
@@ -128,7 +128,7 @@ class Qwen3ASRTextConfig(PretrainedConfig):
     This is the configuration class to store the configuration of a [`Qwen3ASRTextModel`]. It is used to instantiate a
     Qwen3-ASR model according to the specified arguments, defining the model architecture. Instantiating a configuration
     with the defaults will yield a similar configuration to that of
-    Qwen3-ASR-1.7B [Qwen/Qwen3-ASR-1.7B](https://huggingface.co/Qwen/Qwen3-ASR-1.7B)
+    Audar-ASR-V1-Turbo
 
     Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
     documentation from [`PretrainedConfig`] for more information.
@@ -281,10 +281,10 @@ class Qwen3ASRThinkerConfig(PretrainedConfig):
     r"""
     This is the configuration class to store the configuration of a [`Qwen3ASRThinker`]. It is used to instantiate a
     Qwen3-ASR-Thinker model according to the specified arguments, defining the model architecture. Instantiating a
-    configuration with the defaults will yield a similar configuration to that of the thinker component of the Qwen3-Omni
+    configuration with the defaults will yield a similar configuration to that of the decoder component of Audar-ASR-V1-Turbo
     architecture.
 
-    e.g. [Qwen/Qwen3-ASR-1.7B](https://huggingface.co/Qwen/Qwen3-ASR-1.7B)
+    e.g. Audar-ASR-V1-Turbo
 
     Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
     documentation from [`PretrainedConfig`] for more information.
@@ -361,7 +361,7 @@ class Qwen3ASRConfig(PretrainedConfig):
     model according to the specified sub-models configurations, defining the model architecture.
 
     Instantiating a configuration with the defaults will yield a similar configuration to that of the
-    [Qwen/Qwen3-ASR-1.7B](https://huggingface.co/Qwen/Qwen3-ASR-1.7B) architecture.
+    Audar-ASR-V1-Turbo architecture.
 
     Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
     documentation from [`PretrainedConfig`] for more information.