No extra vars call (#3054)

* remove unused reciprocal

* comment

* remove unneeded call to vars

* free speedup

.gitignore

@@ -49,3 +49,4 @@ model.safetensors
 quickstart.py
 .hypothesis
 weights
+*.lprof

test/external/external_test_speed_llama.py

@@ -54,4 +54,5 @@ def run_llama(st, empty_method_cache=True):
     Device[Device.DEFAULT].compiler = backup_compiler
 
 if __name__ == '__main__':
-  unittest.main()
+  TestLLaMASpeed().test_llama_compile()
+  #unittest.main()

tinygrad/lazy.py

@@ -155,37 +155,38 @@ def stride(self, arg:Tuple[int, ...]): return self._view(self.st.stride(arg))
 
 # recursively create a lazyop
 def _recursive_lazyop(buf:LazyBuffer, inputs:List[LazyBuffer], var_vals:Dict[Variable, int], st:ShapeTracker,
-                      realizes:Set[LazyBuffer], first=True, cache=None) -> LazyOp:
-  if cache is None: cache = {}
+                      realizes:Set[LazyBuffer], cache, first=True) -> LazyOp:
   if (buf, st) in cache: return cache[(buf, st)]
   if buf != buf.base:
-    var_vals.update(merge_dicts([var_vals, buf.st.var_vals]))
-    st = buf.st.unbind()+st
+    st = buf.st + st
     buf = buf.base
   # all buffers here are base now
   assert buf.op is not None
 
   # consts are always fused and generated
   if buf.op == LoadOps.CONST:
-    return LazyOp(BufferOps.CONST, (), ConstBuffer(float(buf.arg), buf.dtype, st.simplify()))
+    # TODO: make shapetracker unbind also return var_vals
+    var_vals.update(merge_dicts([var_vals, st.var_vals]))
+    return LazyOp(BufferOps.CONST, (), ConstBuffer(float(buf.arg), buf.dtype, st.simplify().unbind()))
 
   # if we aren't fusing it, it's a load and we add it to the inputs
   if buf.realized or (buf in realizes and not first):
     if buf not in inputs: inputs.append(buf)
-    return LazyOp(BufferOps.LOAD, (), MemBuffer(inputs.index(buf)+1, buf.dtype, st.simplify()))
+    var_vals.update(merge_dicts([var_vals, st.var_vals]))
+    return LazyOp(BufferOps.LOAD, (), MemBuffer(inputs.index(buf)+1, buf.dtype, st.simplify().unbind()))
 
   # if a CONTIGUOUS made it all the way here, just skip it
   if buf.op == LoadOps.CONTIGUOUS:
     assert first
-    return _recursive_lazyop(buf.srcs[0], inputs, var_vals, st, realizes, False, cache)
+    return _recursive_lazyop(buf.srcs[0], inputs, var_vals, st, realizes, cache, False)
 
   # if it's a reduce, we have to change the shapetracker
   if buf.op in ReduceOps:
     assert st.contiguous, "ReduceOps late fusion must be contiguous"
-    st = ShapeTracker.from_shape(buf.srcs[0].shape).unbind()
+    st = ShapeTracker.from_shape(buf.srcs[0].shape)
 
   # otherwise we fuse it like normal
-  cache[(buf, st)] = ret = LazyOp(buf.op, tuple(_recursive_lazyop(x, inputs, var_vals, st, realizes, False, cache) for x in buf.srcs), buf.arg)
+  cache[(buf, st)] = ret = LazyOp(buf.op, tuple(_recursive_lazyop(x, inputs, var_vals, st, realizes, cache, False) for x in buf.srcs), buf.arg)
   return ret
 
 # recursively walk back in the graph to create the schedule
@@ -204,12 +205,11 @@ def _recursive_schedule(out:LazyBuffer, seen:Set[LazyBuffer], realizes:Set[LazyB
   elif out.op == LoadOps.EMPTY:
     op = LazyOp(LoadOps.EMPTY)
   else:
-    output_st = ShapeTracker.from_shape(reduce_for_op[out].shape if out in reduce_for_op else out.shape).unbind()
-    op = _recursive_lazyop(out, inputs, var_vals, output_st, realizes)
-    op = LazyOp(BufferOps.STORE, (op, ), MemBuffer(0, out.dtype, output_st.simplify()))
+    output_st = ShapeTracker.from_shape(reduce_for_op[out].shape if out in reduce_for_op else out.shape)
+    op = _recursive_lazyop(out, inputs, var_vals, output_st, realizes, cache={})
+    op = LazyOp(BufferOps.STORE, (op, ), MemBuffer(0, out.dtype, output_st.simplify().unbind()))
 
-  return flatten(_recursive_schedule(x.base, seen, realizes, reduce_for_op) for x in inputs) + \
-    [ScheduleItem(op, out, tuple(inputs), {k:var_vals[k] for k in op.vars()})]
+  return flatten(_recursive_schedule(x.base, seen, realizes, reduce_for_op) for x in inputs) + [ScheduleItem(op, out, tuple(inputs), var_vals)]
 
 # recursively search the entire graph for all LazyBuffers, insert realizes after expands
 def _recurse_lb(buf:LazyBuffer, realizes:Set[LazyBuffer], allbufs:Dict[LazyBuffer, None],