Merge remote-tracking branch 'refs/remotes/origin/main'

README.md

@@ -101,9 +101,7 @@ pip install -r requirements_nocuda.txt
 
 ```
 cd text-generation-webui
-mkdir repositories
-cd repositories
-git clone https://github.com/turboderp/exllama
+git clone https://github.com/turboderp/exllama repositories/exllama
 ```
 
 #### bitsandbytes on older NVIDIA GPUs

modules/exllama.py

@@ -14,7 +14,7 @@ try:
     from exllama.model import ExLlama, ExLlamaCache, ExLlamaConfig
     from exllama.tokenizer import ExLlamaTokenizer
 except:
-    logger.warning('Exllama module failed to load. Will attempt to load from repositories.')
+    logger.warning('exllama module failed to import. Will attempt to import from repositories/.')
     try:
         from modules.relative_imports import RelativeImport
 
@@ -23,7 +23,10 @@ except:
             from model import ExLlama, ExLlamaCache, ExLlamaConfig
             from tokenizer import ExLlamaTokenizer
     except:
-        logger.error("Could not find repositories/exllama/. Make sure that exllama is cloned inside repositories/ and is up to date.")
+        logger.error(
+            "Could not find repositories/exllama. Please ensure that exllama"
+            " (https://github.com/turboderp/exllama) is cloned inside repositories/ and is up to date."
+        )
         raise
 
 

modules/extensions.py

@@ -27,6 +27,7 @@ def apply_settings(extension, name):
 
 def load_extensions():
     global state, setup_called
+    state = {}
     for i, name in enumerate(shared.args.extensions):
         if name in available_extensions:
             if name != 'api':

requirements.txt

@@ -8,7 +8,7 @@ accelerate==0.22.*
 colorama
 datasets
 einops
-exllamav2==0.0.1
+exllamav2==0.0.2
 markdown
 numpy==1.24
 optimum==1.13.1

requirements_nocuda.txt

@@ -8,7 +8,7 @@ accelerate==0.22.*
 colorama
 datasets
 einops
-exllamav2==0.0.1
+exllamav2==0.0.2
 markdown
 numpy==1.24
 optimum==1.13.1