Spaces:
Running
Running
pass model in the hub
Browse files
app.py
CHANGED
|
@@ -1,13 +1,16 @@
|
|
| 1 |
import gradio as gr
|
| 2 |
from threading import Thread
|
| 3 |
from PIL import Image
|
| 4 |
-
import spaces
|
| 5 |
import moondream as md
|
| 6 |
-
|
| 7 |
-
#subprocess.run('pip install flash-attn --no-build-isolation', env={'FLASH_ATTENTION_SKIP_CUDA_BUILD': "TRUE"}, shell=True)
|
| 8 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
|
| 10 |
-
model = md.vl(model=
|
| 11 |
|
| 12 |
def model_inference(input_dict, history):
|
| 13 |
# Extract image from message if present
|
|
|
|
| 1 |
import gradio as gr
|
| 2 |
from threading import Thread
|
| 3 |
from PIL import Image
|
|
|
|
| 4 |
import moondream as md
|
| 5 |
+
from huggingface_hub import hf_hub_download
|
|
|
|
| 6 |
|
| 7 |
+
# Download model at runtime
|
| 8 |
+
model_path = hf_hub_download(
|
| 9 |
+
repo_id="andito/moondream05",
|
| 10 |
+
filename="moondream-0_5b-int8.mf",
|
| 11 |
+
)
|
| 12 |
|
| 13 |
+
model = md.vl(model=model_path)
|
| 14 |
|
| 15 |
def model_inference(input_dict, history):
|
| 16 |
# Extract image from message if present
|