llama-cpp-python-cuda-gradio

Runtime error

radames commited on Aug 20, 2023

Commit

e1400d9

1 Parent(s): eee4d14

Update app.py

Files changed (1) hide show

app.py CHANGED Viewed

@@ -13,6 +13,7 @@ llm = Llama(
         filename="WizardLM-7B-uncensored.ggmlv3.q4_0.bin",
     ),
     n_ctx=2048,
 )  # download model from hf/ n_ctx=2048 for high ccontext length
 history = []

         filename="WizardLM-7B-uncensored.ggmlv3.q4_0.bin",
     ),
     n_ctx=2048,
+    n_gpu_layers=30
 )  # download model from hf/ n_ctx=2048 for high ccontext length
 history = []