Update llm.py

llm.py

@@ -2,7 +2,7 @@
 import utils; from utils import *
 import os, sys, lzma, json, pprint, time, subprocess
 
-thinker = os.getenv("thinker", "405b")
+thinker = os.getenv("thinker", "70b")
 TEMPERATURE = float(os.getenv("temperature", 0.1)) # 0.0 conservative (good for coding and correct syntax)
 
 LLM_HOST = "gemini"
@@ -82,7 +82,7 @@ elif thinker in "70b|405b":
     # https://docs.together.ai/docs/chat-models#hosted-models
     model = {
         "405b": "meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo  128k", # $3.50 / 1m tokens(*)
-         "70b": "meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo   128k", # $0.88 / 1m tokens(*)
+         "70b": "metameta-llama/Llama-3.3-70B-Instruct-Turbo    128k", # $0.88 / 1m tokens(*)
     }[thinker]
 
     model, CTXLEN = model.strip().split()