Update app/llm.py

app/llm.py

@@ -30,7 +30,7 @@ class ChatModel(BaseModel):
     mirostat_tau: float=4.0
     mirostat_eta: float=1.1
 llm_chat = llama_cpp.Llama.from_pretrained(
-    repo_id="moriire/healthcare-ai-q4_k_m",
+    repo_id="moriire/healthcare-ai-q2_k",
     filename="*.gguf",
     #tokenizer=llama_cpp.llama_tokenizer.LlamaHFTokenizer.from_pretrained("moriire/healthcare-ai-q4_k_m"),
     verbose=False,
@@ -39,8 +39,8 @@ llm_chat = llama_cpp.Llama.from_pretrained(
     #chat_format="llama-2"
 )
 llm_generate = llama_cpp.Llama.from_pretrained(
-    repo_id="moriire/healthcare-ai-q8_0",
-    filename="healthcare-ai-q8_0-unsloth.Q8_0.gguf",
+    repo_id="moriire/healthcare-ai-q2_k",
+    filename="healthcare-ai-q2_k-unsloth.Q2_k.gguf",
     #tokenizer=llama_cpp.llama_tokenizer.LlamaHFTokenizer.from_pretrained("moriire/healthcare-ai-q8_0"),
     verbose=False,
      n_ctx=4096,