Mirrors/text-generation-webui
1
0
Fork 0
mirror of https://github.com/oobabooga/text-generation-webui.git synced 2024-09-20 10:35:10 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

Improve the llama-cpp-python exception messages

Browse source
This commit is contained in:
oobabooga 2024-07-11 13:00:29 -07:00
parent 01e4721da7
commit 512b311137
1 changed files with 14 additions and 4 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

18
modules/llama_cpp_python_hijack.py
View file

@ -13,11 +13,21 @@ imported_module = None
def llama_cpp_lib(): def llama_cpp_lib():
global imported_module global imported_module
def module_to_purpose(module_name):
if module_name == 'llama_cpp':
return 'CPU'
elif module_name == 'llama_cpp_cuda_tensorcores':
return 'tensorcores'
elif module_name == 'llama_cpp_cuda':
return 'default'
return 'unknown'
return_lib = None return_lib = None
if shared.args.cpu: if shared.args.cpu:
if imported_module and imported_module != 'llama_cpp': if imported_module and imported_module != 'llama_cpp':
raise Exception(f"Cannot import 'llama_cpp' because '{imported_module}' is already imported. See issue #1575 in llama-cpp-python. Please restart the server before attempting to use a different version of llama-cpp-python.") raise Exception(f"The {module_to_purpose(imported_module)} version of llama-cpp-python is already loaded. Switching to the CPU version currently requires a server restart.")
try: try:
return_lib = importlib.import_module('llama_cpp') return_lib = importlib.import_module('llama_cpp')
imported_module = 'llama_cpp' imported_module = 'llama_cpp'
@ -26,7 +36,7 @@ def llama_cpp_lib():
if shared.args.tensorcores and return_lib is None: if shared.args.tensorcores and return_lib is None:
if imported_module and imported_module != 'llama_cpp_cuda_tensorcores': if imported_module and imported_module != 'llama_cpp_cuda_tensorcores':
raise Exception(f"Cannot import 'llama_cpp_cuda_tensorcores' because '{imported_module}' is already imported. See issue #1575 in llama-cpp-python. Please restart the server before attempting to use a different version of llama-cpp-python.") raise Exception(f"The {module_to_purpose(imported_module)} version of llama-cpp-python is already loaded. Switching to the tensorcores version currently requires a server restart.")
try: try:
return_lib = importlib.import_module('llama_cpp_cuda_tensorcores') return_lib = importlib.import_module('llama_cpp_cuda_tensorcores')
imported_module = 'llama_cpp_cuda_tensorcores' imported_module = 'llama_cpp_cuda_tensorcores'
@ -35,7 +45,7 @@ def llama_cpp_lib():
if return_lib is None: if return_lib is None:
if imported_module and imported_module != 'llama_cpp_cuda': if imported_module and imported_module != 'llama_cpp_cuda':
raise Exception(f"Cannot import 'llama_cpp_cuda' because '{imported_module}' is already imported. See issue #1575 in llama-cpp-python. Please restart the server before attempting to use a different version of llama-cpp-python.") raise Exception(f"The {module_to_purpose(imported_module)} version of llama-cpp-python is already loaded. Switching to the default version currently requires a server restart.")
try: try:
return_lib = importlib.import_module('llama_cpp_cuda') return_lib = importlib.import_module('llama_cpp_cuda')
imported_module = 'llama_cpp_cuda' imported_module = 'llama_cpp_cuda'
@ -44,7 +54,7 @@ def llama_cpp_lib():
if return_lib is None and not shared.args.cpu: if return_lib is None and not shared.args.cpu:
if imported_module and imported_module != 'llama_cpp': if imported_module and imported_module != 'llama_cpp':
raise Exception(f"Cannot import 'llama_cpp' because '{imported_module}' is already imported. See issue #1575 in llama-cpp-python. Please restart the server before attempting to use a different version of llama-cpp-python.") raise Exception(f"The {module_to_purpose(imported_module)} version of llama-cpp-python is already loaded. Switching to the CPU version currently requires a server restart.")
try: try:
return_lib = importlib.import_module('llama_cpp') return_lib = importlib.import_module('llama_cpp')
imported_module = 'llama_cpp' imported_module = 'llama_cpp'