Fix build issues, update C extension, and enhance tests

Author: SubaashNair

README.md

@@ -6,8 +6,6 @@ OptimRL is a **high-performance reinforcement learning library** that introduces
 ## 🏅 Badges
 
 ![PyPI Version](https://img.shields.io/pypi/v/optimrl)
-<!-- ![PyPI Downloads](https://img.shields.io/pypi/dm/optimrl) -->
-<!-- ![Python Version](https://img.shields.io/pypi/pyversions/optimrl) -->
 ![Python](https://img.shields.io/badge/Python-3.8%2B-blue?logo=python&logoColor=white)
 ![C](https://img.shields.io/badge/C-99-00599C?logo=c&logoColor=white)
 ![NumPy](https://img.shields.io/badge/Library-NumPy-013243?logo=numpy&logoColor=white)

optimrl/c_src/grpo.c

@@ -2,6 +2,21 @@
 #include <stdio.h>
 #include <stdlib.h>
 #include <math.h>
+#define PY_SSIZE_T_CLEAN
+#include <Python.h>
+
+// Example implementation of the PyInit_libgrpo function
+PyMODINIT_FUNC PyInit_libgrpo(void) {
+    static struct PyModuleDef moduledef = {
+        PyModuleDef_HEAD_INIT,
+        "libgrpo",       // Module name
+        "GRPO C extension",  // Module docstring
+        -1,              // Size of per-interpreter state of the module
+        NULL,            // Module methods
+    };
+
+    return PyModule_Create(&moduledef);
+}
 
 // Helper function to compute robust statistics
 void compute_reward_stats(double* rewards, int group_size, double* out_mean, double* out_std) {

optimrl/c_src/libgrpo.dylib

@@ -5,6 +5,11 @@
 import platform
 import shutil
 import versioneer
+import sysconfig
+
+# Get the Python include path dynamically
+python_include_path = sysconfig.get_path('include')
+python_lib_path = sysconfig.get_config_var('LIBDIR')
 
 
 class CustomBuildExt(build_ext):
@@ -77,12 +82,14 @@ def finalize_options(self):
 
 
 # Define the extension module
+
 grpo_module = Extension(
     'optimrl.c_src.libgrpo',
     sources=['optimrl/c_src/grpo.c'],
-    include_dirs=['optimrl/c_src'],
+    include_dirs=['optimrl/c_src',python_include_path],
     libraries=['m'] if platform.system() != 'Windows' else [],
-    extra_compile_args=['-O3', '-fPIC'] if platform.system() != 'Windows' else ['/O2']
+    extra_compile_args=['-O3', '-fPIC'] if platform.system() != 'Windows' else ['/O2'],
+    extra_link_args=[] if platform.system() != 'Windows' else ['/EXPORT:PyInit_libgrpo']
 )
 
 # Read the README file
@@ -110,7 +117,7 @@ def finalize_options(self):
         "torch>=1.8.0"
     ],
     extras_require={
-        'test': ['pytest>=6.0'],
+        'test': ['pytest>=6.0', 'flake8', 'isort', 'black'],
         'dev': ['pytest>=6.0', 'black', 'isort', 'flake8']
     },
     python_requires=">=3.8",