use cuda if available

min_dalle/min_dalle_torch.py

@@ -1,7 +1,7 @@
 import numpy
 from typing import Dict
+from torch import LongTensor, FloatTensor
 import torch
-from torch import Tensor
 torch.no_grad()
 
 from .models.vqgan_detokenizer import VQGanDetokenizer
@@ -15,10 +15,10 @@ from .load_params import (
 
 
 def encode_torch(
-    text_tokens: numpy.ndarray,
+    text_tokens: LongTensor,
     config: dict, 
     params: dict
-) -> Tensor:
+) -> FloatTensor:
     print("loading torch encoder")
     encoder = DalleBartEncoderTorch(
         layer_count = config['encoder_layers'],
@@ -37,20 +37,19 @@ def encode_torch(
     del encoder_params
 
     print("encoding text tokens")
-    text_tokens = torch.tensor(text_tokens).to(torch.long)
     encoder_state = encoder(text_tokens)
     del encoder
     return encoder_state
 
 
 def decode_torch(
-    text_tokens: Tensor,
+    text_tokens: LongTensor,
-    encoder_state: Tensor, 
+    encoder_state: FloatTensor, 
     config: dict,
     seed: int,
     params: dict,
     image_token_count: int
-) -> Tensor:
+) -> LongTensor:
     print("loading torch decoder")
     decoder = DalleBartDecoderTorch(
         image_vocab_size = config['image_vocab_size'],
@@ -86,6 +85,9 @@ def generate_image_tokens_torch(
     params: dict,
     image_token_count: int
 ) -> numpy.ndarray:
+    text_tokens = torch.tensor(text_tokens).to(torch.long)
+    if torch.cuda.is_available(): 
+        text_tokens = text_tokens.cuda()
     encoder_state = encode_torch(
         text_tokens, 
         config,