Spaces:

imperialwool
/

llama-cpp-api

Running

imperialwool commited on Oct 3, 2023

Commit

b8ccea4

•

1 Parent(s): e25a0c4

Update quart_app.py

Files changed (1) hide show

quart_app.py CHANGED Viewed

@@ -9,10 +9,10 @@ app = Quart(__name__)                                              # Quart app
 llm = Llama(model_path="./model.bin")                              # LLaMa model
 llama_model_name = "TheBloke/Llama-2-13B-chat-GGUF"
 translator_tokenizer = M2M100Tokenizer.from_pretrained(            # tokenizer for translator
-    "facebook/m2m100_1.2B", cache_dir="translator/"
 )
 translator_model = M2M100ForConditionalGeneration.from_pretrained( # translator model
-    "facebook/m2m100_1.2B", cache_dir="translator/"
 )
 translator_model.eval()

 llm = Llama(model_path="./model.bin")                              # LLaMa model
 llama_model_name = "TheBloke/Llama-2-13B-chat-GGUF"
 translator_tokenizer = M2M100Tokenizer.from_pretrained(            # tokenizer for translator
+    "facebook/m2m100_418M", cache_dir="translator/"
 )
 translator_model = M2M100ForConditionalGeneration.from_pretrained( # translator model
+    "facebook/m2m100_418M", cache_dir="translator/"
 )
 translator_model.eval()