dev

.gitignore

@@ -1,3 +1,7 @@
+# 
+/model.bin
+
+
 # BASIC PYTHON .GITIGNORE + some for testing
 
 # Byte-compiled / optimized / DLL files

app.py

@@ -15,12 +15,16 @@ from llama_cpp import Llama
 
 # load like this - use tne variable everywhere 
 model_uri_hf=os.getenv("MODEL_URI_HF")
-# DEBUG!
-model_uri_hf="https://huggingface.co/TheBloke/neural-chat-7B-v3-2-GGUF/blob/main/neural-chat-7b-v3-2.Q2_K.gguf"
 # show warning, when empty and briefs description of how to set it 
 # also add link to "how to search" with link to bloke by default + example search link + example full value (mistral base?)
 # info about ram requirements
 
+# DEBUG!
+model_uri_hf="https://huggingface.co/TheBloke/neural-chat-7B-v3-2-GGUF/blob/main/neural-chat-7b-v3-2.Q2_K.gguf"
+model_uri_hf="https://huggingface.co/TheBloke/neural-chat-7B-v3-2-GGUF/resolve/main/neural-chat-7b-v3-2.Q2_K.gguf"
+# maybe use git lfs to dl instead?
+
+
 # Initing things                
 print(f"debug: init model: {model_uri_hf}")