Feat (gpfq): add different way of random projection

src/brevitas/graph/gpfq.py

@@ -2,10 +2,14 @@
 # SPDX-License-Identifier: BSD-3-Clause
 
 from copy import deepcopy
+import math
+from math import pi
 from typing import Callable, List, Optional
 
 import numpy as np
 import torch
+from torch.fft import fft
+from torch.fft import fftn
 import torch.nn as nn
 import unfoldNd
 
@@ -18,7 +22,63 @@
 from brevitas.graph.gpxq import SUPPORTED_CONV_OP
 import brevitas.nn as qnn
 
-TARGET_DIM = 512
+TARGET_DIM = 1024
+USE_MEM_EFF_RP = True
+
+
+def dct_pt(x, norm='ortho'):
+    """
+    Adapted from `https://github.com/zh217/torch-dct`.
+    Discrete Cosine Transform, Type II (a.k.a. the DCT)
+
+    For the meaning of the parameter `norm`, see:
+    https://docs.scipy.org/doc/scipy-0.14.0/reference/generated/scipy.fftpack.dct.html
+
+    :param x: the input signal
+    :param norm: the normalization, None or 'ortho'
+    :return: the DCT-II of the signal over the last dimension
+    """
+    x_shape = x.shape
+    N = x_shape[-1]
+    x = x.reshape(-1, N)
+
+    v = torch.cat([x[:, ::2], x[:, 1::2].flip([1])], dim=1)
+
+    Vc = torch.view_as_real(fft(v, dim=1, norm=norm))
+
+    k = -torch.arange(N, dtype=x.dtype, device=x.device)[None, :] * torch.tensor(pi) / (2 * N)
+    W_r = torch.cos(k)
+    W_i = torch.sin(k)
+
+    V = Vc[:, :, 0] * W_r - Vc[:, :, 1] * W_i
+
+    if norm == 'ortho':
+        V[:, 0] /= torch.sqrt(torch.tensor(N)) * 2
+        V[:, 1:] /= torch.sqrt(torch.tensor(N / 2)) * 2
+
+    V = 2 * V.view(*x_shape)
+
+    return V
+
+
+def random_projection(float_input, quantized_input, target_dim):
+    # our dim to reduce is the last index
+    # creating a diag tensor of 1/-1 entries
+    ones = torch.randint(
+        -1, 1, (quantized_input.size(1), 1), dtype=torch.int8, device=float_input.device)
+    ones[ones == 0] = 1
+    # randomly multiply each row,
+
+    f_out = float_input * ones
+    q_out = quantized_input * ones
+    # apply dite cosine transform
+    f_out = dct_pt(f_out, norm='ortho')
+    q_out = dct_pt(q_out, norm='ortho')
+    # now apply q_R, a random r_n operator, so only keep target_dim of the m rows at random
+    # now apply q_R, a random q_restriction operator, so only keep target_dim of the m rows at random
+    perm = torch.randperm(f_out.size(1))
+    idx = perm[:target_dim].to(float_input.device)
+    return torch.index_select(f_out, dim=1, index=idx), torch.index_select(q_out, dim=1, index=idx)
 
 
 class gpfq_mode(gpxq_mode):
@@ -248,16 +308,20 @@ def single_layer_update(self):
                 weight = weight.transpose(1, 0)  # This performs a view
             weight = weight.flatten(1)
         weight = weight.view(self.groups, -1, weight.shape[-1])  # [Groups, OC/Groups, IC]
-        # use random projection to reduce dimensionality
-        n = self.quantized_input.size(1)
-        # create gaussian random matrix
-        R = torch.normal(mean=0.0, std=1. / n, size=(TARGET_DIM, n), device=dev)
-        self.quantized_input = torch.transpose(self.quantized_input, 1, 2) @ R.T
-        self.float_input = torch.transpose(self.float_input, 1, 2) @ R.T
-        del R
-        # reshape back
-        self.quantized_input = torch.transpose(self.quantized_input, 1, 2).to(dev)
-        self.float_input = torch.transpose(self.float_input, 1, 2).to(dev)
+        if USE_MEM_EFF_RP:
+            self.float_input, self.quantized_input = random_projection(self.float_input, self.quantized_input, TARGET_DIM)
+        else:
+            # use random projection to reduce dimensionality
+            n = self.quantized_input.size(1)
+            # create gaussian random matrix
+            R = torch.normal(mean=0.0, std=1. / math.sqrt(n), size=(TARGET_DIM, n), device=dev)
+            # R = torch.randn(TARGET_DIM, n, device=dev)
+            self.quantized_input = torch.transpose(self.quantized_input, 1, 2) @ R.T
+            self.float_input = torch.transpose(self.float_input, 1, 2) @ R.T
+            del R
+            # reshape back
+            self.quantized_input = torch.transpose(self.quantized_input, 1, 2).to(dev)
+            self.float_input = torch.transpose(self.float_input, 1, 2).to(dev)
         U = torch.zeros(
             weight.shape[0], weight.shape[1], self.float_input.shape[1], device=dev, dtype=dtype)
         # We don't need full Hessian, we just need the diagonal