asr-concurrent-stream 1.0.0__tar.gz

asr_concurrent_stream-1.0.0/PKG-INFO

@@ -0,0 +1,168 @@
+Metadata-Version: 2.4
+Name: asr-concurrent-stream
+Version: 1.0.0
+Summary: Concurrent streaming gRPC server and client for Qwen3-ASR
+Author: Alibaba Qwen Team
+License: Apache-2.0
+Project-URL: Homepage, https://github.com/Qwen/Qwen3-ASR
+Keywords: asr,speech-recognition,grpc,streaming,qwen3
+Requires-Python: >=3.9
+Description-Content-Type: text/markdown
+Requires-Dist: grpcio
+Requires-Dist: numpy
+Requires-Dist: protobuf
+Requires-Dist: vllm
+Requires-Dist: transformers
+Provides-Extra: client
+Requires-Dist: soundfile; extra == "client"
+Requires-Dist: torch; extra == "client"
+Provides-Extra: dev
+Requires-Dist: grpcio-tools; extra == "dev"
+
+# ASR Concurrent Stream Server
+
+Concurrent streaming gRPC server and client for Qwen3-ASR.
+
+## Installation
+
+```bash
+pip install asr-concurrent-stream
+```
+
+For client-side audio loading and VAD segmentation:
+
+```bash
+pip install asr-concurrent-stream[client]
+```
+
+For development (regenerating protobuf code):
+
+```bash
+pip install asr-concurrent-stream[dev]
+```
+
+**Note:** The server requires `vllm` and `transformers` to run the Qwen3-ASR model. These are installed automatically with the package.
+
+## Quick Start
+
+### Start the server
+
+```bash
+asr-concurrent-server
+```
+
+The server reads configuration from environment variables:
+
+| Variable | Default | Description |
+|----------|---------|-------------|
+| `ASR_PORT` | `8000` | gRPC server port |
+| `ASR_MODEL_PATH` | `Qwen/Qwen3-ASR-1.7B` | Model path or HuggingFace ID |
+| `ASR_GPU_MEMORY_UTILIZATION` | `0.30` | GPU memory fraction |
+| `ASR_MAX_MODEL_LEN` | `4096` | Maximum model sequence length |
+| `ASR_MAX_CONCURRENT_STREAMS` | `15` | Maximum concurrent streams |
+| `ASR_MAX_BATCH_SIZE` | `8` | Maximum batch size |
+| `ASR_BATCH_TIMEOUT_MS` | `50` | Batch timeout in milliseconds |
+| `ASR_WORKER_THREADS` | `4` | Number of worker threads |
+| `ASR_HEALTH_PORT` | `8080` | HTTP health check port |
+
+### Run the client
+
+```bash
+asr-concurrent-client --audio path/to/audio.wav --language Italian
+```
+
+## Dependencies
+
+| Package | Purpose |
+|---------|---------|
+| `grpcio` | gRPC framework |
+| `numpy` | Audio array handling |
+| `protobuf` | Protocol buffer serialization |
+| `vllm` | vLLM inference engine |
+| `transformers` | Model processor and tokenizer |
+
+Optional client dependencies: `soundfile`, `torch` (for Silero VAD segmentation).
+
+---
+
+*Docker image for the Qwen3-ASR concurrent streaming gRPC server.*
+
+## Image Composition
+
+| Layer | Detail |
+|-------|--------|
+| **Base OS** | Ubuntu 24.04 LTS |
+| **CUDA** | NVIDIA CUDA 12.8.0 (`nvidia/cuda:12.8.0-runtime-ubuntu24.04`) |
+| **Python** | 3.12 (deadsnakes PPA), installed into `/opt/venv` |
+| **Model** | `Qwen/Qwen3-ASR-1.7B` baked into the image |
+| **Server port** | Container `8002` (map to any host port at runtime) |
+
+## Installation Sequence
+
+1. **System dependencies** — `curl`, `git`, `wget`, `build-essential`, `ffmpeg`, `libsox-*`, `libsndfile1`, `libopus0`, `libffi-dev`.
+2. **Python 3.12** — installed via deadsnakes PPA, venv created at `/opt/venv`, binaries symlinked to `/usr/local/bin`.
+3. **pip** — upgraded to latest `pip`, `setuptools`, `wheel`.
+4. **`qwen-asr[vllm]`** — installs the full Qwen3-ASR stack with vLLM extras (transformers, vllm, torch, accelerate, librosa, soundfile, etc. with pinned versions).
+5. **flash-attention** — prebuilt wheel for CUDA 12.8 + PyTorch 2.9 + Python 3.12.
+6. **Project files** — `ASR_Concurrent_Stream/` copied to `/app/ASR_Concurrent_Stream/`.
+7. **Model download** — `Qwen/Qwen3-ASR-1.7B` downloaded from HuggingFace to `/app/models/Qwen3-ASR-1.7B`.
+8. **Launcher patch** — `run_asr_grpc_server.sh` is updated:
+   - `PORT` → `8002`
+   - `MODEL_PATH` → `/app/models/Qwen3-ASR-1.7B`
+   - `QWEN3_ASR_PATH` → site-packages `qwen_asr` path
+
+## Build & Run
+
+```bash
+# Build
+docker build -t qwen3-asr-server:latest .
+
+# Run (map host port 8002 to container port 8002)
+docker run -d \
+    --name qwen3-asr-server \
+    --gpus all \
+    -p 8002:8002 \
+    --shm-size=16g \
+    --ulimit memlock=-1 \
+    --restart unless-stopped \
+    qwen3-asr-server:latest
+
+# View logs
+docker logs -f qwen3-asr-server
+
+# Stop & remove
+docker stop qwen3-asr-server && docker rm qwen3-asr-server
+```
+
+## Server Configuration
+
+Hard-coded defaults in `server/run_asr_grpc_server.sh`:
+
+| Parameter | Value |
+|-----------|-------|
+| Port | `8002` |
+| Model | `/app/models/Qwen3-ASR-1.7B` |
+| GPU memory utilization | `0.35` |
+| Max model length | `4096` |
+| Max concurrent streams | `10` |
+| Max batch size | `8` |
+| Batch timeout | `50` ms |
+| Worker threads | `4` |
+
+## Project Structure (in image)
+
+```
+/app/
+├── ASR_Concurrent_Stream/
+│   ├── proto/              # gRPC protobuf definitions
+│   ├── server/
+│   │   ├── asr_grpc_server.py
+│   │   ├── _launch_server.py
+│   │   ├── stream_manager.py
+│   │   ├── inference_coordinator.py
+│   │   └── run_asr_grpc_server.sh
+│   └── client/
+│       └── example_client.py
+└── models/
+    └── Qwen3-ASR-1.7B/      # Pre-downloaded model
+```

asr_concurrent_stream-1.0.0/README.md

@@ -0,0 +1,147 @@
+# ASR Concurrent Stream Server
+
+Concurrent streaming gRPC server and client for Qwen3-ASR.
+
+## Installation
+
+```bash
+pip install asr-concurrent-stream
+```
+
+For client-side audio loading and VAD segmentation:
+
+```bash
+pip install asr-concurrent-stream[client]
+```
+
+For development (regenerating protobuf code):
+
+```bash
+pip install asr-concurrent-stream[dev]
+```
+
+**Note:** The server requires `vllm` and `transformers` to run the Qwen3-ASR model. These are installed automatically with the package.
+
+## Quick Start
+
+### Start the server
+
+```bash
+asr-concurrent-server
+```
+
+The server reads configuration from environment variables:
+
+| Variable | Default | Description |
+|----------|---------|-------------|
+| `ASR_PORT` | `8000` | gRPC server port |
+| `ASR_MODEL_PATH` | `Qwen/Qwen3-ASR-1.7B` | Model path or HuggingFace ID |
+| `ASR_GPU_MEMORY_UTILIZATION` | `0.30` | GPU memory fraction |
+| `ASR_MAX_MODEL_LEN` | `4096` | Maximum model sequence length |
+| `ASR_MAX_CONCURRENT_STREAMS` | `15` | Maximum concurrent streams |
+| `ASR_MAX_BATCH_SIZE` | `8` | Maximum batch size |
+| `ASR_BATCH_TIMEOUT_MS` | `50` | Batch timeout in milliseconds |
+| `ASR_WORKER_THREADS` | `4` | Number of worker threads |
+| `ASR_HEALTH_PORT` | `8080` | HTTP health check port |
+
+### Run the client
+
+```bash
+asr-concurrent-client --audio path/to/audio.wav --language Italian
+```
+
+## Dependencies
+
+| Package | Purpose |
+|---------|---------|
+| `grpcio` | gRPC framework |
+| `numpy` | Audio array handling |
+| `protobuf` | Protocol buffer serialization |
+| `vllm` | vLLM inference engine |
+| `transformers` | Model processor and tokenizer |
+
+Optional client dependencies: `soundfile`, `torch` (for Silero VAD segmentation).
+
+---
+
+*Docker image for the Qwen3-ASR concurrent streaming gRPC server.*
+
+## Image Composition
+
+| Layer | Detail |
+|-------|--------|
+| **Base OS** | Ubuntu 24.04 LTS |
+| **CUDA** | NVIDIA CUDA 12.8.0 (`nvidia/cuda:12.8.0-runtime-ubuntu24.04`) |
+| **Python** | 3.12 (deadsnakes PPA), installed into `/opt/venv` |
+| **Model** | `Qwen/Qwen3-ASR-1.7B` baked into the image |
+| **Server port** | Container `8002` (map to any host port at runtime) |
+
+## Installation Sequence
+
+1. **System dependencies** — `curl`, `git`, `wget`, `build-essential`, `ffmpeg`, `libsox-*`, `libsndfile1`, `libopus0`, `libffi-dev`.
+2. **Python 3.12** — installed via deadsnakes PPA, venv created at `/opt/venv`, binaries symlinked to `/usr/local/bin`.
+3. **pip** — upgraded to latest `pip`, `setuptools`, `wheel`.
+4. **`qwen-asr[vllm]`** — installs the full Qwen3-ASR stack with vLLM extras (transformers, vllm, torch, accelerate, librosa, soundfile, etc. with pinned versions).
+5. **flash-attention** — prebuilt wheel for CUDA 12.8 + PyTorch 2.9 + Python 3.12.
+6. **Project files** — `ASR_Concurrent_Stream/` copied to `/app/ASR_Concurrent_Stream/`.
+7. **Model download** — `Qwen/Qwen3-ASR-1.7B` downloaded from HuggingFace to `/app/models/Qwen3-ASR-1.7B`.
+8. **Launcher patch** — `run_asr_grpc_server.sh` is updated:
+   - `PORT` → `8002`
+   - `MODEL_PATH` → `/app/models/Qwen3-ASR-1.7B`
+   - `QWEN3_ASR_PATH` → site-packages `qwen_asr` path
+
+## Build & Run
+
+```bash
+# Build
+docker build -t qwen3-asr-server:latest .
+
+# Run (map host port 8002 to container port 8002)
+docker run -d \
+    --name qwen3-asr-server \
+    --gpus all \
+    -p 8002:8002 \
+    --shm-size=16g \
+    --ulimit memlock=-1 \
+    --restart unless-stopped \
+    qwen3-asr-server:latest
+
+# View logs
+docker logs -f qwen3-asr-server
+
+# Stop & remove
+docker stop qwen3-asr-server && docker rm qwen3-asr-server
+```
+
+## Server Configuration
+
+Hard-coded defaults in `server/run_asr_grpc_server.sh`:
+
+| Parameter | Value |
+|-----------|-------|
+| Port | `8002` |
+| Model | `/app/models/Qwen3-ASR-1.7B` |
+| GPU memory utilization | `0.35` |
+| Max model length | `4096` |
+| Max concurrent streams | `10` |
+| Max batch size | `8` |
+| Batch timeout | `50` ms |
+| Worker threads | `4` |
+
+## Project Structure (in image)
+
+```
+/app/
+├── ASR_Concurrent_Stream/
+│   ├── proto/              # gRPC protobuf definitions
+│   ├── server/
+│   │   ├── asr_grpc_server.py
+│   │   ├── _launch_server.py
+│   │   ├── stream_manager.py
+│   │   ├── inference_coordinator.py
+│   │   └── run_asr_grpc_server.sh
+│   └── client/
+│       └── example_client.py
+└── models/
+    └── Qwen3-ASR-1.7B/      # Pre-downloaded model
+```

asr_concurrent_stream-1.0.0/asr_concurrent_stream.egg-info/PKG-INFO

@@ -0,0 +1,168 @@
+Metadata-Version: 2.4
+Name: asr-concurrent-stream
+Version: 1.0.0
+Summary: Concurrent streaming gRPC server and client for Qwen3-ASR
+Author: Alibaba Qwen Team
+License: Apache-2.0
+Project-URL: Homepage, https://github.com/Qwen/Qwen3-ASR
+Keywords: asr,speech-recognition,grpc,streaming,qwen3
+Requires-Python: >=3.9
+Description-Content-Type: text/markdown
+Requires-Dist: grpcio
+Requires-Dist: numpy
+Requires-Dist: protobuf
+Requires-Dist: vllm
+Requires-Dist: transformers
+Provides-Extra: client
+Requires-Dist: soundfile; extra == "client"
+Requires-Dist: torch; extra == "client"
+Provides-Extra: dev
+Requires-Dist: grpcio-tools; extra == "dev"
+
+# ASR Concurrent Stream Server
+
+Concurrent streaming gRPC server and client for Qwen3-ASR.
+
+## Installation
+
+```bash
+pip install asr-concurrent-stream
+```
+
+For client-side audio loading and VAD segmentation:
+
+```bash
+pip install asr-concurrent-stream[client]
+```
+
+For development (regenerating protobuf code):
+
+```bash
+pip install asr-concurrent-stream[dev]
+```
+
+**Note:** The server requires `vllm` and `transformers` to run the Qwen3-ASR model. These are installed automatically with the package.
+
+## Quick Start
+
+### Start the server
+
+```bash
+asr-concurrent-server
+```
+
+The server reads configuration from environment variables:
+
+| Variable | Default | Description |
+|----------|---------|-------------|
+| `ASR_PORT` | `8000` | gRPC server port |
+| `ASR_MODEL_PATH` | `Qwen/Qwen3-ASR-1.7B` | Model path or HuggingFace ID |
+| `ASR_GPU_MEMORY_UTILIZATION` | `0.30` | GPU memory fraction |
+| `ASR_MAX_MODEL_LEN` | `4096` | Maximum model sequence length |
+| `ASR_MAX_CONCURRENT_STREAMS` | `15` | Maximum concurrent streams |
+| `ASR_MAX_BATCH_SIZE` | `8` | Maximum batch size |
+| `ASR_BATCH_TIMEOUT_MS` | `50` | Batch timeout in milliseconds |
+| `ASR_WORKER_THREADS` | `4` | Number of worker threads |
+| `ASR_HEALTH_PORT` | `8080` | HTTP health check port |
+
+### Run the client
+
+```bash
+asr-concurrent-client --audio path/to/audio.wav --language Italian
+```
+
+## Dependencies
+
+| Package | Purpose |
+|---------|---------|
+| `grpcio` | gRPC framework |
+| `numpy` | Audio array handling |
+| `protobuf` | Protocol buffer serialization |
+| `vllm` | vLLM inference engine |
+| `transformers` | Model processor and tokenizer |
+
+Optional client dependencies: `soundfile`, `torch` (for Silero VAD segmentation).
+
+---
+
+*Docker image for the Qwen3-ASR concurrent streaming gRPC server.*
+
+## Image Composition
+
+| Layer | Detail |
+|-------|--------|
+| **Base OS** | Ubuntu 24.04 LTS |
+| **CUDA** | NVIDIA CUDA 12.8.0 (`nvidia/cuda:12.8.0-runtime-ubuntu24.04`) |
+| **Python** | 3.12 (deadsnakes PPA), installed into `/opt/venv` |
+| **Model** | `Qwen/Qwen3-ASR-1.7B` baked into the image |
+| **Server port** | Container `8002` (map to any host port at runtime) |
+
+## Installation Sequence
+
+1. **System dependencies** — `curl`, `git`, `wget`, `build-essential`, `ffmpeg`, `libsox-*`, `libsndfile1`, `libopus0`, `libffi-dev`.
+2. **Python 3.12** — installed via deadsnakes PPA, venv created at `/opt/venv`, binaries symlinked to `/usr/local/bin`.
+3. **pip** — upgraded to latest `pip`, `setuptools`, `wheel`.
+4. **`qwen-asr[vllm]`** — installs the full Qwen3-ASR stack with vLLM extras (transformers, vllm, torch, accelerate, librosa, soundfile, etc. with pinned versions).
+5. **flash-attention** — prebuilt wheel for CUDA 12.8 + PyTorch 2.9 + Python 3.12.
+6. **Project files** — `ASR_Concurrent_Stream/` copied to `/app/ASR_Concurrent_Stream/`.
+7. **Model download** — `Qwen/Qwen3-ASR-1.7B` downloaded from HuggingFace to `/app/models/Qwen3-ASR-1.7B`.
+8. **Launcher patch** — `run_asr_grpc_server.sh` is updated:
+   - `PORT` → `8002`
+   - `MODEL_PATH` → `/app/models/Qwen3-ASR-1.7B`
+   - `QWEN3_ASR_PATH` → site-packages `qwen_asr` path
+
+## Build & Run
+
+```bash
+# Build
+docker build -t qwen3-asr-server:latest .
+
+# Run (map host port 8002 to container port 8002)
+docker run -d \
+    --name qwen3-asr-server \
+    --gpus all \
+    -p 8002:8002 \
+    --shm-size=16g \
+    --ulimit memlock=-1 \
+    --restart unless-stopped \
+    qwen3-asr-server:latest
+
+# View logs
+docker logs -f qwen3-asr-server
+
+# Stop & remove
+docker stop qwen3-asr-server && docker rm qwen3-asr-server
+```
+
+## Server Configuration
+
+Hard-coded defaults in `server/run_asr_grpc_server.sh`:
+
+| Parameter | Value |
+|-----------|-------|
+| Port | `8002` |
+| Model | `/app/models/Qwen3-ASR-1.7B` |
+| GPU memory utilization | `0.35` |
+| Max model length | `4096` |
+| Max concurrent streams | `10` |
+| Max batch size | `8` |
+| Batch timeout | `50` ms |
+| Worker threads | `4` |
+
+## Project Structure (in image)
+
+```
+/app/
+├── ASR_Concurrent_Stream/
+│   ├── proto/              # gRPC protobuf definitions
+│   ├── server/
+│   │   ├── asr_grpc_server.py
+│   │   ├── _launch_server.py
+│   │   ├── stream_manager.py
+│   │   ├── inference_coordinator.py
+│   │   └── run_asr_grpc_server.sh
+│   └── client/
+│       └── example_client.py
+└── models/
+    └── Qwen3-ASR-1.7B/      # Pre-downloaded model
+```

asr_concurrent_stream-1.0.0/asr_concurrent_stream.egg-info/SOURCES.txt

@@ -0,0 +1,21 @@
+README.md
+pyproject.toml
+asr_concurrent_stream.egg-info/PKG-INFO
+asr_concurrent_stream.egg-info/SOURCES.txt
+asr_concurrent_stream.egg-info/dependency_links.txt
+asr_concurrent_stream.egg-info/entry_points.txt
+asr_concurrent_stream.egg-info/requires.txt
+asr_concurrent_stream.egg-info/top_level.txt
+client/__init__.py
+client/example_client.py
+proto/__init__.py
+proto/asr_pb2.py
+proto/asr_pb2_grpc.py
+server/__init__.py
+server/_launch_server.py
+server/asr_grpc_server.py
+server/asr_model.py
+server/asr_processor.py
+server/asr_utils.py
+server/inference_coordinator.py
+server/stream_manager.py

asr_concurrent_stream-1.0.0/asr_concurrent_stream.egg-info/dependency_links.txt

@@ -0,0 +1 @@
+

asr_concurrent_stream-1.0.0/asr_concurrent_stream.egg-info/entry_points.txt

@@ -0,0 +1,3 @@
+[console_scripts]
+asr-concurrent-client = client.example_client:main
+asr-concurrent-server = server._launch_server:main

asr_concurrent_stream-1.0.0/asr_concurrent_stream.egg-info/requires.txt

@@ -0,0 +1,12 @@
+grpcio
+numpy
+protobuf
+vllm
+transformers
+
+[client]
+soundfile
+torch
+
+[dev]
+grpcio-tools

asr_concurrent_stream-1.0.0/asr_concurrent_stream.egg-info/top_level.txt

@@ -0,0 +1,3 @@
+client
+proto
+server

asr_concurrent_stream-1.0.0/client/__init__.py

@@ -0,0 +1 @@
+# ASR Concurrent Stream Client