switch to big model

modules/app.py

@@ -17,11 +17,18 @@ import json
 # })
 # origins=[f"http://localhost:{PORT}", "https://huggingface.co", "https://hf.space"]
 
-#CUDA = True
-CUDA = False
 PORT = 7860
 VERBOSE = False
 
+BIG_MODEL = True
+
+if big_model:
+    CUDA = True
+    model_name = "facebook/incoder-6B"
+else:
+    CUDA = False
+    model_name = "facebook/incoder-1B"
+
 from fastapi import FastAPI, Request
 from fastapi.staticfiles import StaticFiles
 from fastapi.responses import FileResponse, StreamingResponse
@@ -30,9 +37,9 @@ app.mount("/static", StaticFiles(directory="static"), name="static")
 
 
 print("loading model")
-model = AutoModelForCausalLM.from_pretrained("facebook/incoder-1B")
+model = AutoModelForCausalLM.from_pretrained(model_name)
 print("loading tokenizer")
-tokenizer = AutoTokenizer.from_pretrained("facebook/incoder-1B")
+tokenizer = AutoTokenizer.from_pretrained(model_name)
 print("loading complete")
 
 if CUDA: