cuda device error debug

README.md

@@ -133,4 +133,19 @@ exact_match = 86.91
   biburl    = {https://dblp.org/rec/journals/corr/abs-1810-04805.bib},
   bibsource = {dblp computer science bibliography, https://dblp.org}
 }
-```
+```
+
+
+
+# HF endpoint deployment errors
+
+1
+```
+{'error': 'Body needs to provide a inputs key, recieved: b\'{"question":"What is my name?","context":"My name is Clara and I live in Berkeley."}\''}
+```
+
+2
+```
+{'error': 'Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cpu! (when checking argument for argument index in method wrapper__index_select)'}
+```
+

handler.py

@@ -4,7 +4,7 @@ import torch
 # from scipy.special import softmax
 
 # set device
-device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+# device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 
 # def print_tokens_with_ids(tokenizer, input_ids):
 #     # BERT only needs the token IDs, but for the purpose of inspecting the 
@@ -104,7 +104,8 @@ def get_answer(
 
 class EndpointHandler:
     def __init__(self, path=""):
-        self.model = BertForQuestionAnswering.from_pretrained(path).to(device)
+        # self.model = BertForQuestionAnswering.from_pretrained(path).to(device)
+        self.model = BertForQuestionAnswering.from_pretrained(path)
         self.tokenizer = BertTokenizer.from_pretrained(path)
 
     # def __call__(self, data: Dict[str, Any]):