remove base parameter from fourier_encode function - 0.7.3

README.md

@@ -23,7 +23,6 @@ model = Perceiver(
     input_axis = 2,              # number of axis for input data (2 for images, 3 for video)
     num_freq_bands = 6,          # number of freq bands, with original value (2 * K + 1)
     max_freq = 10.,              # maximum frequency, hyperparameter depending on how fine the data is
-    freq_base = 2,
     depth = 6,                   # depth of net. The shape of the final attention mechanism will be:
                                  #   depth * (cross attention -> self_per_cross_attn * self attention)
     num_latents = 256,           # number of latents, or induced set points, or centroids. different papers giving it different names

perceiver_pytorch/perceiver_pytorch.py

@@ -29,11 +29,11 @@ def cached_fn(*args, _cache = True, **kwargs):
         return cache
     return cached_fn
 
-def fourier_encode(x, max_freq, num_bands = 4, base = 2):
+def fourier_encode(x, max_freq, num_bands = 4):
     x = x.unsqueeze(-1)
     device, dtype, orig_x = x.device, x.dtype, x
 
-    scales = torch.logspace(0., log(max_freq / 2) / log(base), num_bands, base = base, device = device, dtype = dtype)
+    scales = torch.logspace(0., log(max_freq / 2) / log(10), num_bands, device = device, dtype = dtype)
     scales = scales[(*((None,) * (len(x.shape) - 1)), Ellipsis)]
 
     x = x * scales * pi
@@ -128,7 +128,6 @@ def __init__(
         num_freq_bands,
         depth,
         max_freq,
-        freq_base = 2,
         input_channels = 3,
         input_axis = 2,
         num_latents = 512,
@@ -177,7 +176,6 @@ def __init__(
         self.input_axis = input_axis
         self.max_freq = max_freq
         self.num_freq_bands = num_freq_bands
-        self.freq_base = freq_base
 
         self.fourier_encode_data = fourier_encode_data
         fourier_channels = (input_axis * ((num_freq_bands * 2) + 1)) if fourier_encode_data else 0
@@ -231,7 +229,7 @@ def forward(
 
             axis_pos = list(map(lambda size: torch.linspace(-1., 1., steps = size, device = device), axis))
             pos = torch.stack(torch.meshgrid(*axis_pos), dim = -1)
-            enc_pos = fourier_encode(pos, self.max_freq, self.num_freq_bands, base = self.freq_base)
+            enc_pos = fourier_encode(pos, self.max_freq, self.num_freq_bands)
             enc_pos = rearrange(enc_pos, '... n d -> ... (n d)')
             enc_pos = repeat(enc_pos, '... -> b ...', b = b)
 

setup.py

@@ -3,7 +3,7 @@
 setup(
   name = 'perceiver-pytorch',
   packages = find_packages(),
-  version = '0.7.1',
+  version = '0.7.3',
   license='MIT',
   description = 'Perceiver - Pytorch',
   author = 'Phil Wang',