Increase tolerance for fft_test (#2139)

Our Cortex-m55 build is failing for fft_test.cc due to strict tolerance in irfft tests. Increasing tolerance slightly similar to what we did for rfft.

Can test locally with
```make -f tensorflow/lite/micro/tools/make/Makefile -j24 test_kernel_signal_fft_test TARGET=cortex_m_corstone_300 TARGET_ARCH=cortex-m55```

BUG=[287518815](http://b/287518815)

signal/micro/kernels/fft_test.cc

diff --git a/signal/micro/kernels/fft_test.cc b/signal/micro/kernels/fft_test.cc
index e910519..9d6fd25 100644
--- a/signal/micro/kernels/fft_test.cc
+++ b/signal/micro/kernels/fft_test.cc
@@ -428,7 +428,7 @@
                      input_shape, tflite::kIrfftFloatLength512Input,
                      output_shape, tflite::kIrfftFloatLength512Golden,
                      *registration, g_gen_data_fft_length_512_float,
-                     g_gen_data_size_fft_length_512_float, output, 1e-7));
+                     g_gen_data_size_fft_length_512_float, output, 1e-6));
 }
 
 TF_LITE_MICRO_TEST(IrfftTestLength512Int16) {