new neural network arch (faster + better) & minor changes + fixes

Gone is the convolutional neural network, for it turns out not only is
it more difficult to train, but all of the extra information about
board layers didn't make much of a difference at this size.

So cnn1986 has been replaced by nn1986, a standard, two-layer, dense
nn configured as a binary classifier and (mis)used in that capacity.
Note: total number of parameters is unchanged.

HARK: this new nn exposes a bug somewhere in ctak. Run ctlm with
self-play to see the completely borked board state at the end.

1 files changed, 0 insertions, 131 deletions

diff --git a/include/cnn1986.c b/include/cnn1986.c
deleted file mode 100644
index 7c7dbcb..0000000
--- a/include/cnn1986.c
+++ /dev/null
@@ -1,131 +0,0 @@
-/*
-  This file is part of ct.
-
-  This program is free software: you can redistribute it and/or modify
-  it under the terms of the GNU General Public License as published by
-  the Free Software Foundation, either version 3 of the License, or
-  (at your option) any later version.
-
-  This program is distributed in the hope that it will be useful, but
-  WITHOUT ANY WARRANTY; without even the implied warranty of
-  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
-  General Public License for more details.
-
-  You should have received a copy of the GNU General Public License
-  along with ct. If not, see <https://www.gnu.org/licenses/>.
-*/
-
-#include "cnn1986.h"
-#include "weights.h"
-
-// ===================================================================
-// Implementation of a small convolutional neural network
-// ===================================================================
-
-static float cur_board[5*5][KERN_CHAN];
-static float flattened[CONV_NUM+2];
-static float dense1[DENSE1_NUM];
-static float dense2[DENSE2_NUM];
-
-#define RELU(x) ((x) = ((x)<0)?0:(x))
-float cnn1986_evaluate_black_win(void) {
-  /* --------------- *
-   *  Populate input *
-   * --------------- */
-  for (unsigned int y = 0; y < board_size; y++) {
-    for (unsigned int x = 0; x < board_size; x++) {
-      const unsigned int loc = x+y*board_size;
-      const unsigned int count = COUNT_AT(loc);
-      colour_stack_t colour = colours[loc];
-
-      if (count > 0) {
-	float lookup = 0;
-	if (STONE_AT(loc) == STONE_STANDING) {
-	  lookup = (colour & 1) ? +0.25 : -0.25;
-	} else if (STONE_AT(loc) == STONE_CAPSTONE) {
-	  lookup = (colour & 1) ? +1.00 : -1.00;
-	} else {
-	  lookup = (colour & 1) ? +0.50 : -0.50;
-	}
-	cur_board[loc][0] = lookup;
-
-	colour>>=1;
-	for (unsigned int c = 1; c < count && c < KERN_CHAN; c++, colour>>=1) {
-	  cur_board[loc][c] = (colour & 1) ? +0.50 : -0.50;
-	}
-
-	for (unsigned int c = count; c < KERN_CHAN; c++) {
-	  cur_board[loc][c] = 0;
-	}
-      } else {
-	for (unsigned int c = 0; c < KERN_CHAN; c++) {
-	  cur_board[loc][c] = 0;
-	}
-      }
-    }
-  }
-  /* ------------------ *
-   *  Convolution layer *
-   * ------------------ */
-  // for each kernel
-  for (unsigned int kern = 0; kern < KERN_NUM; kern++) {
-    // the stride is 1, march across the board
-    for (unsigned int bx = 0; bx < KERN_OSIZE; bx++) {
-      for (unsigned int by = 0; by < KERN_OSIZE; by++) {
-	flattened[kern+KERN_NUM*(bx+by*KERN_OSIZE)] =
-	  conv2d_biases[kern];
-	// Compute the convolution for this position
-	for (unsigned int ky = 0; ky < KERN_SIZE; ky++) {
-	  for (unsigned int kx = 0; kx < KERN_SIZE; kx++) {
-	    for (unsigned int c = 0; c < KERN_CHAN; c++) {
-	      // Where we are on the board
-	      const unsigned int loc = kx+bx+(ky+by)*board_size;
-	      flattened[kern+KERN_NUM*(bx+by*KERN_OSIZE)]
-		+= cur_board[loc][c]*conv2d_weights[kern][ky][kx][c];
-	    }
-	  }
-	}
-	RELU(flattened[kern+KERN_NUM*(bx+by*KERN_OSIZE)]);
-      }
-    }
-  }
-  // Add input of flat counts
-  flattened[CONV_NUM]   = (float)(white_count & 127)/21.0;
-  flattened[CONV_NUM+1] = (float)(black_count & 127)/21.0;
-  /* ------------------ *
-   *  First dense layer *
-   * ------------------ */
-  for (unsigned int d1 = 0; d1 < DENSE1_NUM; d1++) {
-    dense1[d1] = dense1_biases[d1];
-    for (unsigned int fl = 0; fl < CONV_NUM+2; fl++) {
-      dense1[d1] += flattened[fl]*dense1_weights[d1][fl];
-    }
-    RELU(dense1[d1]);
-  }
-  /* ------------------- *
-   *  Second dense layer *
-   * ------------------- */
-  for (unsigned int d2 = 0; d2 < DENSE2_NUM; d2++) {
-    dense2[d2] = dense2_biases[d2];
-    for (unsigned int d1 = 0; d1 < DENSE1_NUM; d1++) {
-      dense2[d2] += dense1[d1]*dense2_weights[d2][d1];
-    }
-    RELU(dense2[d2]);
-  }
-  /* ------------- *
-   *  Output layer *
-   * ------------- */
-  float output = output_bias;
-  for (unsigned int d2 = 0; d2 < DENSE2_NUM; d2++) {
-    output += dense2[d2]*output_weights[d2];
-  }
-  // 2*(clamped Pade approximant of logistic function) - 1
-  output = (12.0+output+50.0*output/(output*output+10.0))/12.0 - 1.0;
-  if (output > 1.0) {
-    return 1.0;
-  }
-  else if (output < -1.0) {
-    return -1.0;
-  }
-  return output;
-}