TD(lambda), and a big rework

NOTE: 6x6 weights are as-yet untrained

On the performance side:
* switch from linked-list chaining hash table to direct mapping
* cache check_win by hash
* persist transposition tables across plies (this was a big deal and difficult to puzzle out)
* add align directives to the network floats
* prepare_nn_input gained a small LUT to avoid branching (i suppose you might call this 1D embedding)
* apply mtune=native for the native builds

On the correctness side:
* Zobrist has now includes the ply parity
* the network now predicts something in [-1, 1] using tanh (approximated by polynomials)
* the training object is now MSE because we're learning a value estimate
* use .clangd instead of compile_commands.json

On the training side:
* completely forgo human-derived data set
* implement td(lambda), self-play, tournaments

1 files changed, 13 insertions, 0 deletions

diff --git a/.clangd b/.clangd
new file mode 100644
index 0000000..3c5c504
--- /dev/null
+++ b/.clangd
@@ -0,0 +1,13 @@
+# clangd's fallback command runs in the source file's own directory, so the
+# relative -I flags below resolve per file:
+#   -I.          -> include/, for files inside include/
+#   -I../include -> include/, for files inside src/
+CompileFlags:
+  Add:
+    - -std=c99
+    - -I.
+    - -I../include
+    - -D_DEFAULT_SOURCE
+    - -Wall
+    - -Wextra
+    - -Wpedantic