Fixed typo

README.md

@@ -151,7 +151,7 @@ Kera.js can be run in a WebWorker separate from the main thread. Because Keras.j
 
 **WebGL MAX_TEXTURE_SIZE**
 
-In GPU mode, tensor objects are encoded as WebGL textures prior to computations. The size of these tensors are limited by `gl.getParameter(gl.MAX_TEXTURE_SIZE)`, which differs by hardware/platform. See [here](http://webglstats.com/) for typical expected values. The may be an issue in convolution layers after `im2col`. For example, in the Inception V3 network demo, `im2col` in the 1st convolutional layer creates a 22201 x 27 matrix, and 21609 x 288 matrices in the 2nd and 3rd convolutional layers. The size along the first dimension exceeds most `MAX_TEXTURE_SIZE`, 16384, and therefore must be split. Matrix mutiplications are performed with the weights for each split tensor and then combined. In this case, a `weblasTensorsSplit` property is available on the `Tensor` object when `createWeblasTensor()` is called (see `src/Tensor.js`). See `src/layers/convolutional/Convolution2D.js` for an example of its usage.
+In GPU mode, tensor objects are encoded as WebGL textures prior to computations. The size of these tensors are limited by `gl.getParameter(gl.MAX_TEXTURE_SIZE)`, which differs by hardware/platform. See [here](http://webglstats.com/) for typical expected values. The may be an issue in convolution layers after `im2col`. For example, in the Inception V3 network demo, `im2col` in the 1st convolutional layer creates a 22201 x 27 matrix, and 21609 x 288 matrices in the 2nd and 3rd convolutional layers. The size along the first dimension exceeds most `MAX_TEXTURE_SIZE`, 16384, and therefore must be split. Matrix multiplications are performed with the weights for each split tensor and then combined. In this case, a `weblasTensorsSplit` property is available on the `Tensor` object when `createWeblasTensor()` is called (see `src/Tensor.js`). See `src/layers/convolutional/Convolution2D.js` for an example of its usage.
 
 ### Development / Testing