make it compatible to 1 and multiple GPUs
Browse files- handler.py +24 -12
handler.py
CHANGED
@@ -15,22 +15,30 @@ logger.setLevel(logging.DEBUG)
|
|
15 |
|
16 |
# check for GPU
|
17 |
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
18 |
-
|
19 |
|
20 |
class EndpointHandler:
|
21 |
def __init__(self, path=""):
|
22 |
# load the model
|
23 |
self.processor = AutoImageProcessor.from_pretrained("caidas/swin2SR-classical-sr-x2-64")
|
24 |
-
|
25 |
-
|
26 |
-
|
27 |
-
|
28 |
-
|
29 |
-
|
30 |
-
|
31 |
-
|
32 |
-
|
33 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
34 |
def __call__(self, data: Any):
|
35 |
"""
|
36 |
Args:
|
@@ -41,7 +49,11 @@ class EndpointHandler:
|
|
41 |
"""
|
42 |
|
43 |
image = data["inputs"]
|
44 |
-
|
|
|
|
|
|
|
|
|
45 |
|
46 |
try:
|
47 |
with torch.no_grad():
|
|
|
15 |
|
16 |
# check for GPU
|
17 |
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
18 |
+
gpu_count = torch.cuda.device_count()
|
19 |
|
20 |
class EndpointHandler:
|
21 |
def __init__(self, path=""):
|
22 |
# load the model
|
23 |
self.processor = AutoImageProcessor.from_pretrained("caidas/swin2SR-classical-sr-x2-64")
|
24 |
+
|
25 |
+
if(gpu_count > 1):
|
26 |
+
Swin2SRModel._no_split_modules = ["Swin2SREmbeddings", "Swin2SRStage"]
|
27 |
+
Swin2SRForImageSuperResolution._no_split_modules = ["Swin2SREmbeddings", "Swin2SRStage"]
|
28 |
+
model = Swin2SRForImageSuperResolution.from_pretrained("caidas/swin2SR-classical-sr-x2-64", device_map="auto")
|
29 |
+
logger.info(model.hf_device_map)
|
30 |
+
model.hf_device_map["swin2sr.conv_after_body"] = model.hf_device_map["swin2sr.embeddings"]
|
31 |
+
model.hf_device_map["upsample"] = model.hf_device_map["swin2sr.embeddings"]
|
32 |
+
self.model = Swin2SRForImageSuperResolution.from_pretrained("caidas/swin2SR-classical-sr-x2-64", device_map=model.hf_device_map)
|
33 |
+
|
34 |
+
print(subprocess.run(["nvidia-smi"]))
|
35 |
+
|
36 |
+
else:
|
37 |
+
self.model = Swin2SRForImageSuperResolution.from_pretrained("caidas/swin2SR-classical-sr-x2-64")
|
38 |
+
# move model to device
|
39 |
+
self.model.to(device)
|
40 |
+
|
41 |
+
|
42 |
def __call__(self, data: Any):
|
43 |
"""
|
44 |
Args:
|
|
|
49 |
"""
|
50 |
|
51 |
image = data["inputs"]
|
52 |
+
|
53 |
+
if(gpu_count > 1):
|
54 |
+
inputs = self.processor(image, return_tensors="pt")
|
55 |
+
else:
|
56 |
+
inputs = self.processor(image, return_tensors="pt").to(device)
|
57 |
|
58 |
try:
|
59 |
with torch.no_grad():
|