Fix browser init

chat_in_browser.py

@@ -35,7 +35,7 @@ def main():
             except:
                 continue
 
-            if decoded.startswith("System Prompt") and decoded.endswith(": "):
+            if decoded.endswith("Do you want to enter a system prompt? Enter y for yes and anything else for no. \n"):
                 print(f"| {decoded}")
                 proc.stdin.write("\n".encode("utf-8"))
                 proc.stdin.flush()
@@ -93,6 +93,8 @@ def chat():
         model_prefix = "Model: "
         if output.startswith(model_prefix):
             output = output[len(model_prefix) :]
+        else:
+            print("But output is", output)
 
         global convo
 

cli.py

@@ -15,10 +15,7 @@
 from build.utils import allowable_dtype_names, allowable_params_table, get_device_str
 from download import download_and_convert, is_model_downloaded
 
-FORMAT = (
-    "%(levelname)s: %(asctime)-15s: %(filename)s: %(funcName)s: %(module)s: %(message)s"
-)
-logging.basicConfig(filename="/tmp/torchchat.log", level=logging.INFO, format=FORMAT)
+logging.basicConfig(level=logging.INFO,format="%(message)s")
 logger = logging.getLogger(__name__)
 
 default_device = os.getenv("TORCHCHAT_DEVICE", "fast")

generate.py

@@ -752,12 +752,12 @@ def callback(x):
             # Don't continue here.... because we need to report and reset
             # continue
 
-        print(
+        logging.info(
             f"Time for inference {i + 1}: {t:.02f} sec total, {tokens_sec:.02f} tokens/sec"
         )
-        print(f"Bandwidth achieved: {model_size * tokens_sec / 1e9:.02f} GB/s")
+        logging.info(f"Bandwidth achieved: {model_size * tokens_sec / 1e9:.02f} GB/s")
         if i == 0:
-            print(
+            logging.info(
                 f"*** This first iteration will include cold start effects for dynamic import, hardware caches{', JIT compilation' if jit_compile else ''}. ***"
             )
         if start_pos >= max_seq_length: