Qwen3.6-27B iluvatar bi-v100 adaptation

2026-05-21 16:37:24 +08:00
parent fad74b701b
commit 0e89906481
13 changed files with 2283 additions and 11 deletions
--- a/qwen3_6_scripts/patch_transformers_qwen3_5.py
+++ b/qwen3_6_scripts/patch_transformers_qwen3_5.py
@@ -0,0 +1,97 @@
+"""
+Patches transformers 4.55.3 to register the qwen3_5 model type.
+
+Deploy steps on the remote machine:
+  1. cp -r modified_scripts/qwen3_5 /usr/local/lib/python3.10/site-packages/transformers/models/qwen3_5
+  2. python3 modified_scripts/patch_transformers_qwen3_5.py
+
+Target: pip-installed transformers at /usr/local/lib/python3.10/site-packages/transformers/
+(Not the corex pre-installed path at /usr/local/corex/lib64/python3/dist-packages/)
+"""
+
+import sys
+
+TRANSFORMERS_ROOT = "/usr/local/lib/python3.10/site-packages/transformers"
+AUTO_CONFIG = f"{TRANSFORMERS_ROOT}/models/auto/configuration_auto.py"
+MODELS_INIT = f"{TRANSFORMERS_ROOT}/models/__init__.py"
+
+
+def patch_file(path, replacements):
+    with open(path, "r") as f:
+        content = f.read()
+
+    patched = False
+    for old, new in replacements:
+        if new in content:
+            print(f"  [skip] already patched: {repr(new[:60])}")
+            continue
+        if old not in content:
+            print(f"  [warn] anchor not found: {repr(old[:60])}")
+            continue
+        content = content.replace(old, new, 1)
+        patched = True
+        print(f"  [ok]   inserted after: {repr(old[:60])}")
+
+    if patched:
+        with open(path, "w") as f:
+            f.write(content)
+
+
+def main():
+    print(f"=== Patching {AUTO_CONFIG} ===")
+    patch_file(AUTO_CONFIG, [
+        # CONFIG_MAPPING_NAMES: insert qwen3_5 right after qwen3
+        (
+            '("qwen3", "Qwen3Config"),',
+            '("qwen3", "Qwen3Config"),\n        ("qwen3_5", "Qwen3_5Config"),',
+        ),
+        # Some versions don't have trailing comma — handle that too
+        (
+            '("qwen3", "Qwen3Config")\n',
+            '("qwen3", "Qwen3Config"),\n        ("qwen3_5", "Qwen3_5Config"),\n',
+        ),
+        # MODEL_NAMES_MAPPING (model_type -> human readable name, used by docstring generator)
+        (
+            '("qwen3", "Qwen3"),',
+            '("qwen3", "Qwen3"),\n        ("qwen3_5", "Qwen3_5"),',
+        ),
+        (
+            '("qwen3", "Qwen3")\n',
+            '("qwen3", "Qwen3"),\n        ("qwen3_5", "Qwen3_5"),\n',
+        ),
+    ])
+
+    print(f"\n=== Patching {MODELS_INIT} ===")
+    patch_file(MODELS_INIT, [
+        (
+            "from .qwen3 import *\n",
+            "from .qwen3 import *\n    from .qwen3_5 import *\n",
+        ),
+    ])
+
+    # Verification
+    print("\n=== Verification ===")
+    try:
+        import importlib.util, types
+
+        # Quick smoke-test: import the config class directly
+        spec = importlib.util.spec_from_file_location(
+            "configuration_qwen3_5",
+            f"{TRANSFORMERS_ROOT}/models/qwen3_5/configuration_qwen3_5.py",
+        )
+        mod = importlib.util.module_from_spec(spec)
+        # Provide minimal parent package stubs so relative imports resolve
+        pkg = types.ModuleType("transformers")
+        pkg.__path__ = [TRANSFORMERS_ROOT]
+        sys.modules.setdefault("transformers", pkg)
+        spec.loader.exec_module(mod)
+        cfg = mod.Qwen3_5Config()
+        print(f"  Qwen3_5Config() smoke-test OK  (model_type={cfg.model_type})")
+    except Exception as e:
+        print(f"  [warn] smoke-test failed (may be fine at runtime): {e}")
+
+    print("\nDone.")
+
+
+if __name__ == "__main__":
+    main()