Upload 19 files

.gitattributes

@@ -33,3 +33,9 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+bin/aarch64/test_sr filter=lfs diff=lfs merge=lfs -text
+bin/x86_64/test_sr filter=lfs diff=lfs merge=lfs -text
+edsr_x2_128.axmodel filter=lfs diff=lfs merge=lfs -text
+lib/aarch64/libsr.so filter=lfs diff=lfs merge=lfs -text
+lib/gradio_demo.jpg filter=lfs diff=lfs merge=lfs -text
+lib/x86_64/libsr.so filter=lfs diff=lfs merge=lfs -text

bin/aarch64/test_sr

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be1bf087a6018e31f5b4c17c02d1887d98f2e6cd9d3e292f0e48c22939b62e6e
+size 5706312

bin/x86_64/test_sr

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:85c62a92ae91224cf4332d905b1f9bfcd04914c757372d05a7ef5e378c324f98
+size 4947088

edsr_x2_128.axmodel

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:77997756612342ed8e422be4c47bbe9bf04511bdc5483d3d953d188e5018ddb9
+size 1469507

include/ax_devices.h

@@ -0,0 +1,123 @@
+#ifndef __AX_DEVICES_H__
+#define __AX_DEVICES_H__
+
+#if defined(__cplusplus)
+extern "C"
+{
+#endif
+#define AX_DEVICES_COUNT 16
+#define AX_VERSION_LEN 32
+
+    typedef enum
+    {
+        ax_dev_errcode_failed = -1,
+        ax_dev_errcode_success = 0,
+
+        ax_dev_errcode_sysinit_failed,
+        ax_dev_errcode_sysdeinit_failed,
+        ax_dev_errcode_axcl_sysinit_failed,
+        ax_dev_errcode_axcl_sysdeinit_failed,
+    } ax_dev_errcode_e;
+
+    typedef enum
+    {
+        unknown_device = 0,
+        host_device = 1,
+        axcl_device = 2
+    } ax_devive_e;
+
+    typedef struct
+    {
+        struct
+        {
+            char available;
+            char version[AX_VERSION_LEN];
+            struct
+            {
+                int remain;
+                int total;
+            } mem_info;
+        } host;
+
+        struct
+        {
+            char host_version[AX_VERSION_LEN];
+            char dev_version[AX_VERSION_LEN];
+            unsigned char count;
+            struct
+            {
+                int temp;
+                int cpu_usage;
+                int npu_usage;
+                struct
+                {
+                    int remain;
+                    int total;
+                } mem_info;
+            } devices_info[AX_DEVICES_COUNT];
+
+        } devices;
+    } ax_devices_t;
+
+    int ax_dev_enum_devices(ax_devices_t *devices);
+    int ax_dev_sys_init(ax_devive_e dev_type, char devid);
+    int ax_dev_sys_deinit(ax_devive_e dev_type, char devid);
+
+    /**
+     * @brief 当设备初始化后，申请设备内存
+     * 当devid>=0，则虚拟地址为host地址（用作对称），物理地址为设备地址
+     * 当devid=-1，则虚拟地址和物理地址都为CMM地址
+     *
+     * @param devid 设备id
+     * @param size 内存大小
+     * @param pVirAddr 虚拟地址
+     * @param phyAddr 物理地址
+     * @return int
+     */
+    int ax_dev_malloc(char devid, int size, void **pVirAddr, unsigned long long int *phyAddr);
+
+    /**
+     * @brief 当设备初始化后，释放设备内存
+     *
+     * @param devid 设备id
+     * @param pVirAddr 虚拟地址
+     * @param phyAddr 物理地址
+     * @return int
+     */
+    int ax_dev_free(char devid, void *pVirAddr, unsigned long long int phyAddr);
+
+    /**
+     * @brief 当设备初始化后，将内存从源地址复制到目标地址(注意，加速卡内存拷贝不会将ax_dev_malloc申请的host内存进行同步，需要用户自己进行同步)
+     * src和dst都是host内存，则执行memcpy
+     * src为host内存，而dst为设备内存，则执行 host2dev memcpy
+     * dst为host内存，而src为设备内存，则执行 dev2host memcpy
+     * src和dst都是设备内存，则执行dev2dev memcpy（仅支持同设备id内 memcpy）
+     * 
+     *
+     * @param devid 设备id
+     * @param pSrcVirAddr 源虚拟地址
+     * @param SrcphyAddr 源物理地址，如果为0，则认为为host的内存
+     * @param pDstVirAddr 目标虚拟地址
+     * @param DstphyAddr 目标物理地址，如果为0，则认为为host的内存
+     * @param size 内存大小
+     * @return int
+     */
+    int ax_dev_memcpy(char devid, void *pSrcVirAddr, unsigned long long int SrcphyAddr, void *pDstVirAddr, unsigned long long int DstphyAddr, int size);
+
+    /**
+     * @brief 当设备初始化后，将内存设置为指定值
+     *
+     * @param devid 设备id
+     * @param pVirAddr 虚拟地址
+     * @param phyAddr 物理地址
+     * @param value 值
+     * @param size 内存大小
+     * @return int
+     */
+    int ax_dev_memset(char devid, void *pVirAddr, unsigned long long int phyAddr, int value, int size);
+
+#if defined(__cplusplus)
+}
+#endif
+
+#endif // __AX_DEVICES_H__

include/ax_sr.h

@@ -0,0 +1,38 @@
+#ifndef _AX_SR_H_
+#define _AX_SR_H_
+
+#if defined(__cplusplus)
+extern "C"
+{
+#endif
+#include "ax_devices.h"
+
+    typedef struct
+    {
+        ax_devive_e dev_type; // Device type
+        char devid;           // axcl device ID
+        char model_path[256];
+    } ax_sr_init_param_t;
+
+    typedef struct
+    {
+        int width;
+        int height;
+
+        void *pVirAddr;
+    } ax_sr_image_t;
+
+    typedef void *ax_sr_handle_t;
+
+    int ax_sr_init(ax_sr_init_param_t *param, ax_sr_handle_t *handle);
+    int ax_sr_deinit(ax_sr_handle_t handle);
+
+    int ax_sr_get_scale(ax_sr_handle_t handle);
+
+    int ax_sr_run(ax_sr_handle_t handle, ax_sr_image_t *input, ax_sr_image_t *output);
+
+#if defined(__cplusplus)
+}
+#endif
+
+#endif

lib/aarch64/libsr.so

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:374146847484b9c421e5d85688122c93f5389edccaf195851d3a7d15c3c0fd8c
+size 3274264

lib/example.py

@@ -0,0 +1,41 @@
+from pysr import SR
+from pyaxdev import enum_devices, sys_init, sys_deinit, AxDeviceType
+import cv2
+import argparse
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser()
+    parser.add_argument('--model', type=str, default='edsr_x2_small_1.axmodel')
+    parser.add_argument('--image', type=str)
+    args = parser.parse_args()
+
+    devices_info = enum_devices()
+    dev_type = None
+    devid = None
+    print("可用设备:", devices_info)
+    if devices_info['host']['available']:
+        print("host device available")
+        sys_init(AxDeviceType.host_device, -1)
+        dev_type = AxDeviceType.host_device
+        devid = -1
+    elif devices_info['devices']['count'] > 0:
+        print("axcl device available, use device-0")
+        sys_init(AxDeviceType.axcl_device, 0)
+        dev_type = AxDeviceType.axcl_device
+        devid = 0
+    else:
+        raise Exception("No available device")
+    
+    sr = SR({
+            'dev_type': dev_type,
+            'devid': devid,
+            'model_path': args.model
+        })
+    image = cv2.imread(args.image)
+    sr_image = sr(image)
+    cv2.imwrite('sr_image.jpg', sr_image)
+    
+    if devices_info['host']['available']:
+        sys_deinit(AxDeviceType.host_device, -1)
+    elif devices_info['devices']['count'] > 0:
+        sys_deinit(AxDeviceType.axcl_device, 0)

lib/gradio_example.py

@@ -0,0 +1,67 @@
+import gradio as gr
+from pysr import SR
+from pyaxdev import enum_devices, sys_init, sys_deinit, AxDeviceType
+import argparse
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser()
+    parser.add_argument('--model', type=str, default='edsr_x2_small_1.axmodel')
+    args = parser.parse_args()
+
+    devices_info = enum_devices()
+    dev_type = None
+    devid = None
+    print("可用设备:", devices_info)
+    if devices_info['host']['available']:
+        print("host device available")
+        sys_init(AxDeviceType.host_device, -1)
+        dev_type = AxDeviceType.host_device
+        devid = -1
+    elif devices_info['devices']['count'] > 0:
+        print("axcl device available, use device-0")
+        sys_init(AxDeviceType.axcl_device, 0)
+        dev_type = AxDeviceType.axcl_device
+        devid = 0
+    else:
+        raise Exception("No available device")
+    
+    sr = SR({
+            'dev_type': dev_type,
+            'devid': devid,
+            'model_path': args.model
+        })
+
+
+ 
+    def run_super_resolution(image):
+        # image 是 numpy 数组 (H, W, 3)，BGR 格式
+        if image is None:
+            return None
+        sr_image = sr(image)
+        return (image, sr_image)  # 返回两个图，给ImageComparator显示
+
+    with gr.Blocks() as demo:
+        gr.Markdown("## 图像超分对比 Demo")
+        with gr.Row():
+            with gr.Column(scale=1):
+                input_image = gr.Image(label="上传图片", type="numpy")
+                run_button = gr.Button("运行超分")
+            with gr.Column(scale=4):
+                output_comparator = gr.ImageSlider(height=800)
+
+        run_button.click(fn=run_super_resolution, inputs=input_image, outputs=output_comparator)
+
+    
+    # 启动
+    ip = "0.0.0.0"
+    demo.launch(server_name=ip, server_port=7860)
+    
+    
+
+    import atexit
+    if devices_info['host']['available']:
+        atexit.register(lambda: sys_deinit(AxDeviceType.host_device, -1))
+    elif devices_info['devices']['count'] > 0:
+        atexit.register(lambda: sys_deinit(AxDeviceType.axcl_device, 0))
+    
+    

lib/pyaxdev.py

@@ -0,0 +1,150 @@
+import ctypes
+import os
+import platform
+
+def check_error(code: int) -> None:
+    if code != 0:
+        raise Exception(f"API错误: {code}")
+
+base_dir = os.path.dirname(__file__)
+arch = platform.machine()
+
+if arch == 'x86_64':
+    arch_dir = 'x86_64'
+elif arch in ('aarch64', 'arm64'):
+    arch_dir = 'aarch64'
+else:
+    raise RuntimeError(f"Unsupported architecture: {arch}")
+
+so_name = "libsr.so"
+lib_paths = [
+    os.path.join(base_dir, arch_dir, so_name),
+    os.path.join(base_dir, so_name),
+    os.path.join("../build/", so_name),
+    os.path.join("../build_aarch64/", so_name)
+]
+
+last_error = None
+diagnostic_shown = set()
+
+for lib_path in lib_paths:
+    try:
+        print(f"Trying to load: {lib_path}")
+        _lib = ctypes.CDLL(lib_path)
+        print(f"✅ Successfully loaded: {lib_path}")
+        break
+    except OSError as e:
+        last_error = e
+        err_str = str(e)
+        print(f"\n❌ Failed to load: {lib_path}")
+        print(f"   {err_str}")
+
+        # Only show GLIBCXX tip once
+        if "GLIBCXX" in err_str and "not found" in err_str:
+            if "missing_glibcxx" not in diagnostic_shown:
+                diagnostic_shown.add("missing_glibcxx")
+                print("🔍 Detected missing GLIBCXX version in libstdc++.so.6")
+                print("💡 This usually happens when your environment (like Conda) uses an older libstdc++")
+                print(f"👉 Try running with system libstdc++ preloaded:")
+                print(f"   export LD_PRELOAD=/usr/lib/{arch_dir}-linux-gnu/libstdc++.so.6\n")
+        elif "No such file" in err_str:
+            if "file_not_found" not in diagnostic_shown:
+                diagnostic_shown.add("file_not_found")
+                print("🔍 File not found. Please verify that libclip.so exists and the path is correct.\n")
+        elif "wrong ELF class" in err_str:
+            if "elf_mismatch" not in diagnostic_shown:
+                diagnostic_shown.add("elf_mismatch")
+                print("🔍 ELF class mismatch — likely due to architecture conflict (e.g., loading x86_64 .so on aarch64).")
+                print(f"👉 Run `file {lib_path}` to verify the binary architecture.\n")
+        else:
+            if "generic_error" not in diagnostic_shown:
+                diagnostic_shown.add("generic_error")
+                print("📎 Tip: Use `ldd` to inspect missing dependencies:")
+                print(f"   ldd {lib_path}\n")
+else:
+    raise RuntimeError(f"\n❗ Failed to load libclip.so.\nLast error:\n{last_error}")
+
+
+# 定义枚举类型
+class AxDeviceType(ctypes.c_int):
+    unknown_device = 0
+    host_device = 1
+    axcl_device = 2
+
+# 定义结构体
+class AxMemInfo(ctypes.Structure):
+    _fields_ = [
+        ('remain', ctypes.c_int),
+        ('total', ctypes.c_int)
+    ]
+
+class AxHostInfo(ctypes.Structure):
+    _fields_ = [
+        ('available', ctypes.c_char),
+        ('version', ctypes.c_char * 32),
+        ('mem_info', AxMemInfo)
+    ]
+
+class AxDeviceInfo(ctypes.Structure):
+    _fields_ = [
+        ('temp', ctypes.c_int),
+        ('cpu_usage', ctypes.c_int),
+        ('npu_usage', ctypes.c_int),
+        ('mem_info', AxMemInfo)
+    ]
+
+class AxDevices(ctypes.Structure):
+    _fields_ = [
+        ('host', AxHostInfo),
+        ('host_version', ctypes.c_char * 32),
+        ('dev_version', ctypes.c_char * 32),
+        ('count', ctypes.c_ubyte),
+        ('devices_info', AxDeviceInfo * 16)
+    ]
+
+
+_lib.ax_dev_enum_devices.argtypes = [ctypes.POINTER(AxDevices)]
+_lib.ax_dev_enum_devices.restype = ctypes.c_int
+
+_lib.ax_dev_sys_init.argtypes = [AxDeviceType, ctypes.c_char]
+_lib.ax_dev_sys_init.restype = ctypes.c_int
+
+_lib.ax_dev_sys_deinit.argtypes = [AxDeviceType, ctypes.c_char]
+_lib.ax_dev_sys_deinit.restype = ctypes.c_int
+
+def enum_devices() -> dict:
+    devices = AxDevices()
+    check_error(_lib.ax_dev_enum_devices(ctypes.byref(devices)))
+    
+    return {
+        'host': {
+            'available': bool(devices.host.available[0]),
+            'version': devices.host.version.decode('utf-8'),
+            'mem_info': {
+                'remain': devices.host.mem_info.remain,
+                'total': devices.host.mem_info.total
+            }
+        },
+        'devices': {
+            'host_version': devices.host_version.decode('utf-8'),
+            'dev_version': devices.dev_version.decode('utf-8'),
+            'count': devices.count,
+            'devices_info': [{
+                'temp': dev.temp,
+                'cpu_usage': dev.cpu_usage,
+                'npu_usage': dev.npu_usage,
+                'mem_info': {
+                    'remain': dev.mem_info.remain,
+                    'total': dev.mem_info.total
+                }
+            } for dev in devices.devices_info[:devices.count]]
+        }
+    }
+
+
+def sys_init(dev_type: AxDeviceType = AxDeviceType.axcl_device, devid: int = 0) -> None:
+    check_error(_lib.ax_dev_sys_init(dev_type, devid))
+
+
+def sys_deinit(dev_type: AxDeviceType = AxDeviceType.axcl_device, devid: int = 0) -> None:
+    check_error(_lib.ax_dev_sys_deinit(dev_type, devid))

lib/pysr.py

@@ -0,0 +1,62 @@
+import ctypes
+import numpy as np
+from pyaxdev import _lib, AxDeviceType, AxDevices, check_error
+
+
+class SRInit(ctypes.Structure):
+    _fields_ = [
+        ('dev_type', AxDeviceType),
+        ('devid', ctypes.c_char),
+        ('model_path', ctypes.c_char * 256)
+    ]
+
+class SRImage(ctypes.Structure):
+    _fields_ = [
+        ('width', ctypes.c_int),
+        ('height', ctypes.c_int),
+        ('pVirAddr', ctypes.POINTER(ctypes.c_ubyte))
+    ]
+
+
+_lib.ax_sr_init.argtypes = [ctypes.POINTER(SRInit), ctypes.POINTER(ctypes.c_void_p)]
+_lib.ax_sr_init.restype = ctypes.c_int
+
+_lib.ax_sr_deinit.argtypes = [ctypes.c_void_p]
+_lib.ax_sr_deinit.restype = ctypes.c_int
+
+_lib.ax_sr_run.argtypes = [ctypes.c_void_p, ctypes.POINTER(SRImage), ctypes.POINTER(SRImage)]
+_lib.ax_sr_run.restype = ctypes.c_int
+
+
+class SR:
+    def __init__(self, init_info: dict):
+        self.handle = None
+        self.init_info = SRInit()
+        
+        # 设置初始化参数
+        self.init_info.dev_type = init_info.get('dev_type', AxDeviceType.axcl_device)
+        self.init_info.devid = init_info.get('devid', 0)
+        setattr(self.init_info, 'model_path', init_info['model_path'].encode('utf-8'))
+
+        handle = ctypes.c_void_p()
+        check_error(_lib.ax_sr_init(ctypes.byref(self.init_info), ctypes.byref(handle)))
+        self.handle = handle
+
+    def __del__(self):
+        if self.handle:
+            _lib.ax_sr_deinit(self.handle)
+
+    def __call__(self, image_data: np.ndarray) -> None:
+        image = SRImage()
+        image.width = image_data.shape[1]
+        image.height = image_data.shape[0]
+        image.pVirAddr = ctypes.cast(image_data.ctypes.data, ctypes.POINTER(ctypes.c_ubyte))
+            
+        np_sr_image = np.zeros((image.height*2, image.width*2, 3), dtype=np.uint8)
+        sr_image = SRImage()
+        sr_image.width = image.width*2
+        sr_image.height = image.height*2
+        sr_image.pVirAddr = ctypes.cast(np_sr_image.ctypes.data, ctypes.POINTER(ctypes.c_ubyte))
+        
+        check_error(_lib.ax_sr_run(self.handle, ctypes.byref(image), ctypes.byref(sr_image)))
+        return np_sr_image

lib/requirements.txt

@@ -0,0 +1 @@
+gradio

lib/x86_64/libsr.so

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e97d880bb5af1d11bf9cf34f25fb0d346b7aaa4a53df22a80835fa09391eecff
+size 3730592