Add loguru, pynvml

app.py

@@ -3,57 +3,83 @@ from __future__ import annotations
 from typing import Iterable
 
 import gradio as gr
-import torch
+import pynvml
+
+# import torch
 from ctransformers import AutoModelForCausalLM
 from gradio.themes.base import Base
 from gradio.themes.utils import colors, fonts, sizes
 from huggingface_hub import hf_hub_download  # snapshot_download,
+from loguru import logger
 
-repo_id = 'TheBloke/openbuddy-mistral-7B-v13-GGUF'
-filename = 'openbuddy-mistral-7b-v13.Q4_K_S.gguf'  # 4.17G
+repo_id = "TheBloke/openbuddy-mistral-7B-v13-GGUF"
+filename = "openbuddy-mistral-7b-v13.Q4_K_S.gguf"  # 4.17G
 
+logger.debug("Start dl")
 model_path = hf_hub_download(repo_id=repo_id, filename=filename, revision="main")
+logger.debug("Done dl")
 
 # Set gpu_layers to the number of layers to offload to GPU. Set to 0 if no GPU acceleration is available on your system.
 # model = AutoModelForCausalLM.from_pretrained("TheBloke/openbuddy-mistral-7B-v13-GGUF", model_file="openbuddy-mistral-7b-v13.Q4_K_S.gguf", model_type="mistral", gpu_layers=0)
 
-if torch.cuda.is_available():
+has_cuda = False
+try:
+    pynvml.nvmlInit()
+    has_cuda = True
+    logger.debug("has cuda")
+except pynvml.nvml.NVMLError_LibraryNotFound:
+    logger.debug("no cuda")
+
+# if torch.cuda.is_available():
+if has_cuda:
     gpu_layers = 50  # set to what you like for GPU
 else:
     gpu_layers = 0
-model = AutoModelForCausalLM.from_pretrained(model_path, model_type="mistral", gpu_layers=gpu_layers)
 
-ins = '''[INST] <<FRIDAY>>
+logger.debug("Start loading the model")
+model = AutoModelForCausalLM.from_pretrained(
+    model_path, model_type="mistral", gpu_layers=gpu_layers
+)
+logger.debug("Done loading the model")
+
+ins = """[INST] <<FRIDAY>>
 Remember that your English name is "Shi-Ci" and your name in Chinese is "兮辞". You are a helpful, respectful and honest assistant. Always answer as helpfully as possible, while being safe.  Your answers should not include any harmful, unethical, racist, sexist, toxic, dangerous, or illegal content. Please ensure that your responses are socially unbiased and positive in nature.
 If a question does not make any sense, or is not factually coherent, explain why instead of answering something not correct. If you don't know the answer to a question, please don't share false information.
 <</FRIDAY>>
 {} [/INST]
-'''
-
+"""
 
 theme = gr.themes.Monochrome(
     primary_hue="indigo",
     secondary_hue="blue",
     neutral_hue="slate",
     radius_size=gr.themes.sizes.radius_sm,
-    font=[gr.themes.GoogleFont("Open Sans"), "ui-sans-serif", "system-ui", "sans-serif"],
+    font=[
+        gr.themes.GoogleFont("Open Sans"),
+        "ui-sans-serif",
+        "system-ui",
+        "sans-serif",
+    ],
 )
+
+
 def response(question):
     res = model(ins.format(question))
     yield res
 
 
-examples = [
-    "Hello!"
-]
+examples = ["Hello!"]
+
 
 def process_example(args):
     for x in response(args):
         pass
     return x
 
+
 css = ".generating {visibility: hidden}"
 
+
 # Based on the gradio theming guide and borrowed from https://huggingface.co/spaces/shivi/dolly-v2-demo
 class SeafoamCustom(Base):
     def __init__(
@@ -117,9 +143,12 @@ with gr.Blocks(theme=seafoam, analytics_enabled=False, css=css) as demo:
         )
 
         with gr.Row():
-
             with gr.Column(scale=3):
-                instruction = gr.Textbox(placeholder="Enter your question here", label="Question", elem_id="q-input")
+                instruction = gr.Textbox(
+                    placeholder="Enter your question here",
+                    label="Question",
+                    elem_id="q-input",
+                )
 
                 with gr.Box():
                     gr.Markdown("**Answer**")
@@ -133,9 +162,7 @@ with gr.Blocks(theme=seafoam, analytics_enabled=False, css=css) as demo:
                     outputs=[output],
                 )
 
-
-
     submit.click(response, inputs=[instruction], outputs=[output])
     instruction.submit(response, inputs=[instruction], outputs=[output])
 
-demo.queue(concurrency_count=1).launch(debug=False,share=True)
+demo.queue(concurrency_count=1).launch(debug=False, share=True)

requirements.txt

@@ -1,3 +1,5 @@
 ctransformers
 gradio
-huggingface-hub
+huggingface-hub
+loguru
+pynvml