set fallback llama model to use smallest one also

d0b7154f6366100331b30acddfb9c455cdccd8c8 · 2024-10-31 17:15:05 -0400 · Mukund Mauji

Files touched

M exo/api/chatgpt_api.py

Diff

commit d0b7154f6366100331b30acddfb9c455cdccd8c8
Author: Mukund Mauji <mauji.mukund@yahoo.ca>
Date:   Thu Oct 31 17:15:05 2024 -0400

    set fallback llama model to use smallest one also
---
 exo/api/chatgpt_api.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/exo/api/chatgpt_api.py b/exo/api/chatgpt_api.py
index 9ea6b4a9..e596f036 100644
--- a/exo/api/chatgpt_api.py
+++ b/exo/api/chatgpt_api.py
@@ -231,7 +231,7 @@ class ChatGPTAPI:
     stream = data.get("stream", False)
     chat_request = parse_chat_request(data, self.default_model)
     if chat_request.model and chat_request.model.startswith("gpt-"):  # to be compatible with ChatGPT tools, point all gpt- model requests to llama instead
-      chat_request.model = self.default_model if self.default_model.startswith("llama") else "llama-3.1-8b"
+      chat_request.model = self.default_model if self.default_model.startswith("llama") else "llama-3.2-1b"
     if not chat_request.model or chat_request.model not in model_base_shards:
       if DEBUG >= 1: print(f"Invalid model: {chat_request.model}. Supported: {list(model_base_shards.keys())}. Defaulting to {self.default_model}")
       chat_request.model = self.default_model

← 9987ce84 use the smallest possible model as the default_model · back to Exo · use KVCache instead of RotatingKVCache 661035f0 →