Spaces:
Runtime error
Runtime error
cpu
Browse files- inference.py +2 -1
inference.py
CHANGED
|
@@ -29,7 +29,8 @@ def load_model(
|
|
| 29 |
|
| 30 |
model = AutoModelForCausalLM.from_pretrained(
|
| 31 |
pretrained_model_name_or_path=model_name,
|
| 32 |
-
torch_dtype=torch.float32 # Use float32 for CPU
|
|
|
|
| 33 |
).to(device)
|
| 34 |
|
| 35 |
model.eval() # Set model to evaluation mode
|
|
|
|
| 29 |
|
| 30 |
model = AutoModelForCausalLM.from_pretrained(
|
| 31 |
pretrained_model_name_or_path=model_name,
|
| 32 |
+
torch_dtype=torch.float32, # Use float32 for CPU
|
| 33 |
+
low_cpu_mem_usage=True # Helps with memory efficiency
|
| 34 |
).to(device)
|
| 35 |
|
| 36 |
model.eval() # Set model to evaluation mode
|