diff --git a/notebooks/FCLayer_graph.onnx b/notebooks/FCLayer_graph.onnx
index 641ef704256c0c7af863cafc79b3f6af57d0abd5..4cd3fa2804a9960ed405da87d02b0b15b5909bf4 100644
Binary files a/notebooks/FCLayer_graph.onnx and b/notebooks/FCLayer_graph.onnx differ
diff --git a/notebooks/FINN-CodeGenerationAndCompilation.ipynb b/notebooks/FINN-CodeGenerationAndCompilation.ipynb
index f33d4baf724753edb2418a02b47f8bb9aac44af9..e81bc5a364b67caf6cac6be78317b8e306277cdc 100644
--- a/notebooks/FINN-CodeGenerationAndCompilation.ipynb
+++ b/notebooks/FINN-CodeGenerationAndCompilation.ipynb
@@ -13,7 +13,7 @@
},
{
"cell_type": "code",
- "execution_count": 12,
+ "execution_count": 3,
"metadata": {},
"outputs": [],
"source": [
@@ -45,7 +45,7 @@
},
{
"cell_type": "code",
- "execution_count": 13,
+ "execution_count": 4,
"metadata": {},
"outputs": [],
"source": [
@@ -64,7 +64,7 @@
},
{
"cell_type": "code",
- "execution_count": 14,
+ "execution_count": 5,
"metadata": {},
"outputs": [],
"source": [
@@ -73,7 +73,6 @@
"mh = 8\n",
"pe = 4\n",
"simd = 4\n",
- "wmem = mw * mh // (pe * simd)\n",
"nf = mh // pe\n",
"sf = mw // simd\n"
]
@@ -87,7 +86,7 @@
},
{
"cell_type": "code",
- "execution_count": 15,
+ "execution_count": 6,
"metadata": {},
"outputs": [],
"source": [
@@ -110,7 +109,7 @@
},
{
"cell_type": "code",
- "execution_count": 16,
+ "execution_count": 7,
"metadata": {},
"outputs": [],
"source": [
@@ -127,8 +126,8 @@
" MH=mh,\n",
" SIMD=simd,\n",
" PE=pe,\n",
- " WMEM=wmem,\n",
- " TMEM=0,\n",
+ " noActivation=1,\n",
+ " binaryXnorMode=1,\n",
" inputDataType=idt.name,\n",
" weightDataType=wdt.name,\n",
" outputDataType=odt.name,\n",
@@ -144,7 +143,7 @@
},
{
"cell_type": "code",
- "execution_count": 17,
+ "execution_count": 8,
"metadata": {},
"outputs": [],
"source": [
@@ -162,7 +161,7 @@
},
{
"cell_type": "code",
- "execution_count": 18,
+ "execution_count": 9,
"metadata": {},
"outputs": [],
"source": [
@@ -185,7 +184,7 @@
},
{
"cell_type": "code",
- "execution_count": 19,
+ "execution_count": 10,
"metadata": {},
"outputs": [],
"source": [
@@ -194,15 +193,13 @@
},
{
"cell_type": "code",
- "execution_count": 20,
+ "execution_count": 11,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
- "\n",
- "Stopping http://0.0.0.0:8081\n",
"Serving 'FCLayer_graph.onnx' at http://0.0.0.0:8081\n"
]
}
@@ -214,7 +211,7 @@
},
{
"cell_type": "code",
- "execution_count": 21,
+ "execution_count": 12,
"metadata": {},
"outputs": [
{
@@ -247,7 +244,7 @@
},
{
"cell_type": "code",
- "execution_count": 22,
+ "execution_count": 13,
"metadata": {},
"outputs": [
{
@@ -261,6 +258,8 @@
" for node in model.graph.node:\n",
" if node.domain == \"finn\":\n",
" backend_attribute = get_by_name(node.attribute, \"backend\")\n",
+ " if backend_attribute is None:\n",
+ " continue\n",
" backend_value = backend_attribute.s.decode(\"UTF-8\")\n",
" if backend_value == \"fpgadataflow\":\n",
" _codegen_single_node(node, model)\n",
@@ -283,7 +282,7 @@
},
{
"cell_type": "code",
- "execution_count": 23,
+ "execution_count": 14,
"metadata": {},
"outputs": [
{
diff --git a/notebooks/FINN-CustomOps.ipynb b/notebooks/FINN-CustomOps.ipynb
index 1f60546e342564dfbd6b3edf5472f0ba63ddbb0a..6d0e3b33ba1bf2787498ef7e85449e5db2d1e272 100644
--- a/notebooks/FINN-CustomOps.ipynb
+++ b/notebooks/FINN-CustomOps.ipynb
@@ -18,7 +18,7 @@
},
{
"cell_type": "code",
- "execution_count": 8,
+ "execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
@@ -56,12 +56,12 @@
" MH=mh,\n",
" SIMD=simd,\n",
" PE=pe,\n",
- " WMEM=wmem,\n",
- " TMEM=tmem,\n",
- " inputDataType=FINN-DataType,\n",
- " weightDataType=FINN-DataType,\n",
- " outputDataType=FINN-DataType,\n",
+ " inputDataType=<FINN DataType>,\n",
+ " weightDataType=<FINN DataType>,\n",
+ " outputDataType=<FINN DataType>,\n",
" ActVal=actval,\n",
+ " binaryXnorMode=<0/1>,\n",
+ " noActivation=<0/1>\n",
")`"
]
},
@@ -76,12 +76,12 @@
"cell_type": "markdown",
"metadata": {},
"source": [
- "<font size=\"3\">Custom Ops are represented in Finn as ONNX nodes on the one hand and by a CustomOp class on the other hand. This allows easier access to the different attributes and introduces special custom op functions. See below for the standard CustomOp class.</font>"
+ "<font size=\"3\">Custom Ops are represented in FINN as ONNX nodes on the one hand and by a CustomOp class on the other hand. This allows easier access to the different attributes and introduces special custom op functions. See below for the standard CustomOp class.</font>"
]
},
{
"cell_type": "code",
- "execution_count": 3,
+ "execution_count": 2,
"metadata": {},
"outputs": [
{
@@ -188,7 +188,7 @@
},
{
"cell_type": "code",
- "execution_count": 9,
+ "execution_count": 3,
"metadata": {},
"outputs": [
{
@@ -399,7 +399,7 @@
},
{
"cell_type": "code",
- "execution_count": 10,
+ "execution_count": 4,
"metadata": {
"scrolled": true
},
@@ -431,15 +431,13 @@
" self.get_nodeattr(\"SIMD\"),\n",
" export_wdt.get_hls_datatype_str(),\n",
" self.get_nodeattr(\"PE\"),\n",
- " self.get_nodeattr(\"WMEM\"),\n",
+ " self.calc_wmem(),\n",
" )\n",
" )\n",
" else:\n",
" f_weights.write(\n",
" \"static BinaryWeights<{},{},{}> weights = \".format(\n",
- " self.get_nodeattr(\"SIMD\"),\n",
- " self.get_nodeattr(\"PE\"),\n",
- " self.get_nodeattr(\"WMEM\"),\n",
+ " self.get_nodeattr(\"SIMD\"), self.get_nodeattr(\"PE\"), self.calc_wmem()\n",
" )\n",
" )\n",
" f_weights.write(weight_hls_code)\n",
@@ -453,6 +451,12 @@
" # use UINT32 threshold export for bipolar times bipolar\n",
" inp_is_bipolar = self.get_input_datatype() == DataType.BIPOLAR\n",
" wt_is_bipolar = self.get_weight_datatype() == DataType.BIPOLAR\n",
+ " # reinterpret inp/wt as bipolar if bin_xnor_mode is iset\n",
+ " inp_is_binary = self.get_input_datatype() == DataType.BINARY\n",
+ " wt_is_binary = self.get_weight_datatype() == DataType.BINARY\n",
+ " bin_xnor_mode = self.get_nodeattr(\"binaryXnorMode\") == 1\n",
+ " inp_is_bipolar = inp_is_bipolar or (inp_is_binary and bin_xnor_mode)\n",
+ " wt_is_bipolar = wt_is_bipolar or (wt_is_binary and bin_xnor_mode)\n",
" if inp_is_bipolar and wt_is_bipolar:\n",
" tdt = DataType.UINT32\n",
" thresholds_hls_code = numpy_to_hls_code(\n",
@@ -470,7 +474,7 @@
" f_thresh.write(\n",
" \"static ThresholdsActivation<{},{},{},{},{},{},{}> threshs \\\n",
" = \".format(\n",
- " self.get_nodeattr(\"TMEM\"),\n",
+ " self.calc_tmem(),\n",
" self.get_nodeattr(\"PE\"),\n",
" threshold_tensor.shape[-1],\n",
" tdt_hls,\n",