Remove RecurrentReuseFactor from the Altera backend (#1581)

* Remove RecurrentReuseFactor from the Altera backend

On Altera, the SimpleRNN, LSTM and GRU initializers overwrote the
user's RecurrentReuseFactor with the layer's ReuseFactor, so the
setting that config_from_keras_model wrote into every config had no
effect.

Stop registering the setting, matching what #1577 does for Vivado and
Vitis, and warn when a config still sets it. The internal
recurrent_reuse_factor attribute is still set from ReuseFactor.

Fixes #1580

* Move _warn_removed_reuse_factors into FPGABackend

The Vivado and Altera backends had identical copies of the helper; keep a
single one in the shared FPGABackend class.

Author: Anandb71

docs/ir/attributes.rst

@@ -2191,7 +2191,7 @@ Backend-specific attributes
 
   * The number of times each multiplier is used by controlling the amount of pipelining/unrolling. A lower number results in more parallelism and lower latency at the expense of the resources used. Reuse factor = 1 corresponds to all multiplications executed in parallel, and hence, the lowest possible latency.
 
-  * Available in: Quartus, Catapult, Altera
+  * Available in: Quartus, Catapult
 
 LSTM
 ====
@@ -2295,7 +2295,7 @@ Backend-specific attributes
 
   * The number of times each multiplier is used by controlling the amount of pipelining/unrolling. A lower number results in more parallelism and lower latency at the expense of the resources used. Reuse factor = 1 corresponds to all multiplications executed in parallel, and hence, the lowest possible latency.
 
-  * Available in: Quartus, Catapult, Altera
+  * Available in: Quartus, Catapult
 
 GRU
 ===
@@ -2401,7 +2401,7 @@ Backend-specific attributes
 
   * The number of times each multiplier is used by controlling the amount of pipelining/unrolling. A lower number results in more parallelism and lower latency at the expense of the resources used. Reuse factor = 1 corresponds to all multiplications executed in parallel, and hence, the lowest possible latency.
 
-  * Available in: Quartus, Catapult, Altera
+  * Available in: Quartus, Catapult
 
 Bidirectional
 =============

hls4ml/backends/altera/altera_backend.py

@@ -37,7 +37,7 @@ def __init__(self):
         self._register_flows()
 
     def _register_layer_attributes(self):
-        # Add RNN-specific recurrent_reuse_factor attribute
+        # Add RNN-specific table_size and table attributes
         rnn_layers = [
             SimpleRNN,
             LSTM,
@@ -46,7 +46,6 @@ def _register_layer_attributes(self):
 
         for layer in rnn_layers:
             attrs = self.attribute_map.get(layer, [])
-            attrs.append(ConfigurableAttribute('recurrent_reuse_factor', default=1, description=descriptions.reuse_factor))
             attrs.append(ConfigurableAttribute('table_size', default=1024, description=descriptions.table_size))
             attrs.append(TypeAttribute('table', default=FixedPrecisionType(18, 8), description=descriptions.table_type))
             self.attribute_map[layer] = attrs
@@ -280,6 +279,7 @@ def init_embed(self, layer):
 
     @layer_optimizer(GRU)
     def init_gru(self, layer):
+        self._warn_removed_reuse_factors(layer, ['recurrent_reuse_factor'])
         reuse_factor = layer.model.config.get_reuse_factor(layer)
         layer.set_attr('recurrent_reuse_factor', reuse_factor)
 
@@ -367,6 +367,7 @@ def init_conv2d(self, layer):
 
     @layer_optimizer(LSTM)
     def init_lstm(self, layer):
+        self._warn_removed_reuse_factors(layer, ['recurrent_reuse_factor'])
         reuse_factor = layer.model.config.get_reuse_factor(layer)
         layer.set_attr('recurrent_reuse_factor', reuse_factor)
 
@@ -412,6 +413,7 @@ def init_lstm(self, layer):
 
     @layer_optimizer(SimpleRNN)
     def init_simple_rnn(self, layer):
+        self._warn_removed_reuse_factors(layer, ['recurrent_reuse_factor'])
         reuse_factor = layer.model.config.get_reuse_factor(layer)
         layer.set_attr('recurrent_reuse_factor', reuse_factor)
 

hls4ml/backends/fpga/fpga_backend.py

@@ -3,6 +3,7 @@
 import subprocess
 from bisect import bisect_left
 from collections.abc import Iterable
+from warnings import warn
 
 import numpy as np
 
@@ -50,6 +51,7 @@
     XnorPrecisionType,
 )
 from hls4ml.utils import attribute_descriptions as descriptions
+from hls4ml.utils.string_utils import convert_to_pascal_case
 from hls4ml.writer import get_writer
 
 
@@ -299,6 +301,16 @@ def get_closest_reuse_factor(self, valid_rf, chosen_rf):
         else:
             return before
 
+    def _warn_removed_reuse_factors(self, layer, attributes):
+        """Warn about reuse factor settings that are no longer used but are still in the config."""
+
+        for attribute in attributes:
+            if attribute in layer.attributes:
+                warn(
+                    f'{convert_to_pascal_case(attribute)} of layer "{layer.name}" ({layer.class_name}) is no longer used. '
+                    'The ReuseFactor of the layer applies to all of its multiplications.'
+                )
+
     def set_closest_reuse_factor(self, layer, n_in, n_out, attribute='reuse_factor', include_max_rf=True):
         assert attribute is not None, 'Reuse factor attribute cannot be None'
 

hls4ml/backends/vivado/vivado_backend.py

@@ -38,7 +38,6 @@
 from hls4ml.report import parse_vivado_report
 from hls4ml.utils import attribute_descriptions as descriptions
 from hls4ml.utils.einsum_utils import parse_einsum
-from hls4ml.utils.string_utils import convert_to_pascal_case
 
 
 class VivadoBackend(FPGABackend):
@@ -688,16 +687,6 @@ def _validate_recurrent_activations(self, layer, attributes):
                     f'({layer.class_name}) is not supported. Use "linear", "relu", "sigmoid" or "tanh".'
                 )
 
-    def _warn_removed_reuse_factors(self, layer, attributes):
-        """Warn about reuse factor settings that are no longer used but are still in the config."""
-
-        for attribute in attributes:
-            if attribute in layer.attributes:
-                warn(
-                    f'{convert_to_pascal_case(attribute)} of layer "{layer.name}" ({layer.class_name}) is no longer used. '
-                    'The ReuseFactor of the layer applies to all of its multiplications.'
-                )
-
     def _set_recurrent_cell_schedule(self, layer, n_in, n_state, prefix=''):
         """Set the schedule of the cell of the Resource strategy (see nnet_recurrent.h) from the reuse factor, after
         replacing a reuse factor the cell cannot build with the closest one it can."""

test/pytest/test_rnn.py

@@ -872,3 +872,35 @@ def test_bidirectional_gru_reset_after_accuracy(test_case_id, reset_after, stati
     keras_prediction = model.predict(X)
     hls_prediction = hls_model.predict(X).reshape(keras_prediction.shape)
     np.testing.assert_allclose(hls_prediction, keras_prediction, rtol=0.0, atol=5e-2)
+
+
+@pytest.mark.parametrize('rnn_layer', [SimpleRNN, LSTM, GRU], ids=['SimpleRNN', 'LSTM', 'GRU'])
+def test_altera_recurrent_reuse_factor_removed(test_case_id, rnn_layer):
+    """Altera ignores RecurrentReuseFactor and warns; the recurrent multiplication uses ReuseFactor."""
+    model = Sequential()
+    model.add(Input(shape=(8, 6)))
+    model.add(rnn_layer(16, name='rnn'))
+
+    config = hls4ml.utils.config_from_keras_model(model, granularity='name', backend='Altera')
+    assert 'RecurrentReuseFactor' not in config['LayerName']['rnn']
+
+    config['LayerName']['rnn']['ReuseFactor'] = 8
+    config['LayerName']['rnn']['RecurrentReuseFactor'] = 4
+    output_dir = str(test_root_path / test_case_id)
+    with pytest.warns(UserWarning, match='RecurrentReuseFactor'):
+        hls_model = hls4ml.converters.convert_from_keras_model(
+            model, hls_config=config, output_dir=output_dir, backend='Altera'
+        )
+
+    assert hls_model.graph['rnn'].get_attr('recurrent_reuse_factor') == 8
+
+    del config['LayerName']['rnn']['RecurrentReuseFactor']
+    with warnings.catch_warnings(record=True) as caught:
+        warnings.simplefilter('always')
+        hls4ml.converters.convert_from_keras_model(
+            model,
+            hls_config=config,
+            output_dir=str(test_root_path / f'{test_case_id}_nowarn'),
+            backend='Altera',
+        )
+    assert not any('RecurrentReuseFactor' in str(w.message) for w in caught)