PQ with pytorch (facebookresearch#4116)

Summary:
Pull Request resolved: facebookresearch#4116

This diff implements Product Quantization using Pytorch only.

Reviewed By: mdouze

Differential Revision: D67766798

fbshipit-source-id: fe2d44a674fc2056f7e2082e9765052c98fdc8f8

Author: mlomeli1

contrib/torch/clustering.py

@@ -13,6 +13,7 @@
 # the kmeans can produce both torch and numpy centroids
 from faiss.contrib.clustering import kmeans
 
+
 class DatasetAssign:
     """Wrapper for a tensor that offers a function to assign the vectors
     to centroids. All other implementations offer the same interface"""

contrib/torch/quantization.py

@@ -7,33 +7,47 @@
 This contrib module contains Pytorch code for quantization.
 """
 
-import numpy as np
 import torch
 import faiss
-
-from faiss.contrib import torch_utils
+import math
+from faiss.contrib.torch import clustering
+# the kmeans can produce both torch and numpy centroids
 
 
 class Quantizer:
 
     def __init__(self, d, code_size):
+        """
+        d: dimension of vectors
+        code_size: nb of bytes of the code (per vector)
+        """
         self.d = d
         self.code_size = code_size
 
     def train(self, x):
+        """
+        takes a n-by-d array and peforms training
+        """
         pass
 
     def encode(self, x):
+        """
+        takes a n-by-d float array, encodes to an n-by-code_size uint8 array
+        """
         pass
 
-    def decode(self, x):
+    def decode(self, codes):
+        """
+        takes a n-by-code_size uint8 array, returns a n-by-d array
+        """
         pass
 
 
 class VectorQuantizer(Quantizer):
 
     def __init__(self, d, k):
-        code_size = int(torch.ceil(torch.log2(k) / 8))
+
+        code_size = int(math.ceil(torch.log2(k) / 8))
         Quantizer.__init__(d, code_size)
         self.k = k
 
@@ -42,12 +56,41 @@ def train(self, x):
 
 
 class ProductQuantizer(Quantizer):
-
     def __init__(self, d, M, nbits):
-        code_size = int(torch.ceil(M * nbits / 8))
-        Quantizer.__init__(d, code_size)
+        """ M: number of subvectors, d%M == 0
+        nbits: number of bits that each vector is encoded into
+        """
+        assert d % M == 0
+        assert nbits == 8  # todo: implement other nbits values
+        code_size = int(math.ceil(M * nbits / 8))
+        Quantizer.__init__(self, d, code_size)
         self.M = M
         self.nbits = nbits
+        self.code_size = code_size
 
     def train(self, x):
-        pass
+        nc = 2 ** self.nbits
+        sd = self.d // self.M
+        dev = x.device
+        dtype = x.dtype
+        self.codebook = torch.zeros((self.M, nc, sd), device=dev, dtype=dtype)
+        for m in range(self.M):
+            xsub = x[:, m * self.d // self.M: (m + 1) * self.d // self.M]
+            data = clustering.DatasetAssign(xsub.contiguous())
+            self.codebook[m] = clustering.kmeans(2 ** self.nbits, data)
+
+    def encode(self, x):
+        codes = torch.zeros((x.shape[0], self.code_size), dtype=torch.uint8)
+        for m in range(self.M):
+            xsub = x[:, m * self.d // self.M:(m + 1) * self.d // self.M]
+            _, I = faiss.knn(xsub.contiguous(), self.codebook[m], 1)
+            codes[:, m] = I.ravel()
+        return codes
+
+    def decode(self, codes):
+        idxs = [codes[:, m].long() for m in range(self.M)]
+        vectors = [self.codebook[m, idxs[m], :] for m in range(self.M)]
+        stacked_vectors = torch.stack(vectors, dim=1)
+        cbd = self.codebook.shape[-1]
+        x_rec = stacked_vectors.reshape(-1, cbd * self.M)
+        return x_rec

tests/torch_test_contrib.py

@@ -4,13 +4,14 @@
 # LICENSE file in the root directory of this source tree.
 
 import torch  # usort: skip
-import unittest   # usort: skip
-import numpy as np   # usort: skip
+import unittest  # usort: skip
+import numpy as np  # usort: skip
 
-import faiss   # usort: skip
+import faiss  # usort: skip
 import faiss.contrib.torch_utils  # usort: skip
 from faiss.contrib import datasets
-from faiss.contrib.torch import clustering
+from faiss.contrib.torch import clustering, quantization
+
 
 
 
@@ -400,3 +401,27 @@ def test_python_kmeans(self):
         # 33498.332 33380.477
         # print(err, err2)        1/0
         self.assertLess(err2, err * 1.1)
+
+
+class TestQuantization(unittest.TestCase):
+    def test_python_product_quantization(self):
+        """ Test the python implementation of product quantization """
+        d = 64
+        n = 10000
+        cs = 4
+        nbits = 8
+        M = 4
+        x = np.random.random(size=(n, d)).astype('float32')
+        pq = faiss.ProductQuantizer(d, cs, nbits)
+        pq.train(x)
+        codes = pq.compute_codes(x)
+        x2 = pq.decode(codes)
+        diff = ((x - x2)**2).sum()
+        # vs pure pytorch impl
+        xt = torch.from_numpy(x)
+        my_pq = quantization.ProductQuantizer(d, M, nbits)
+        my_pq.train(xt)
+        my_codes = my_pq.encode(xt)
+        xt2 = my_pq.decode(my_codes)
+        my_diff = ((xt - xt2)**2).sum()
+        self.assertLess(abs(diff - my_diff), 100)