Upgrading faster_whisperer to 1.1.1 and a bunch of other fixes to the original worker

Dockerfile

@@ -1,5 +1,5 @@
 # Use specific version of nvidia cuda image
-FROM nvidia/cuda:11.7.1-cudnn8-runtime-ubuntu20.04
+FROM nvidia/cuda:12.3.2-cudnn9-runtime-ubuntu22.04
 
 # Remove any third-party apt sources to avoid issues with expiring keys.
 RUN rm -f /etc/apt/sources.list.d/*.list

builder/fetch_models.py

@@ -1,7 +1,7 @@
 from concurrent.futures import ThreadPoolExecutor
 from faster_whisper import WhisperModel
 
-model_names = ["tiny", "base", "small", "medium", "large-v1", "large-v2", "large-v3"]
+model_names = ["tiny"] #, "base", "small", "medium", "large-v1", "large-v2", "large-v3"]
 
 
 def load_model(selected_model):

builder/requirements.txt

@@ -1,3 +1,6 @@
-runpod~=1.7.0
+#runpod~=1.7.0
+git+https://github.com/nerdylive123/runpod-python.git@fix-validation-bug
 
-faster-whisper==0.10.0
+faster-whisper==1.1.1
+
+hf_xet==0.1.4

src/predict.py

@@ -30,7 +30,7 @@ def load_model(self, model_name):
 
     def setup(self):
         """Load the model into memory to make running multiple predictions efficient"""
-        model_names = ["tiny", "base", "small", "medium", "large-v1", "large-v2", "large-v3"]
+        model_names = ["tiny"] # , "base", "small", "medium", "large-v1", "large-v2", "large-v3"]
         with ThreadPoolExecutor() as executor:
             for model_name, model in executor.map(self.load_model, model_names):
                 if model_name is not None:

src/rp_handler.py

@@ -10,7 +10,8 @@
 
 from rp_schema import INPUT_VALIDATIONS
 from runpod.serverless.utils import download_files_from_urls, rp_cleanup, rp_debugger
-from runpod.serverless.utils.rp_validator import validate
+from runpod.serverless.utils.rp_validator import validate 
+
 import runpod
 import predict