From dbef56df477cb8a80ccafa183144c6123f577f30 Mon Sep 17 00:00:00 2001
From: nihuini <nihuini@tencent.com>
Date: Tue, 30 Apr 2019 11:57:10 +0800
Subject: [PATCH] write graph

---
 tools/pytorch/pytorch2ncnn.cpp | 184 +++++++++++++++++++++++++++++++--
 1 file changed, 178 insertions(+), 6 deletions(-)

diff --git a/tools/pytorch/pytorch2ncnn.cpp b/tools/pytorch/pytorch2ncnn.cpp
index 1d51d493e..8c1c57f16 100644
--- a/tools/pytorch/pytorch2ncnn.cpp
+++ b/tools/pytorch/pytorch2ncnn.cpp
@@ -17,6 +17,7 @@
 #include <string.h>
 
 #include <map>
+#include <set>
 #include <string>
 #include <vector>
 
@@ -439,6 +440,8 @@ class PyTorchNode
 {
 public:
     std::string op;
+    std::string name;
+    std::vector<std::string> inputs;
     std::vector<std::string> outputs;
     std::vector<std::string> args;
 };
@@ -563,6 +566,8 @@ static int read_code(const std::string& code, std::vector<PyTorchNode>& nodes)
 {
     nodes.clear();
 
+    int internal_unknown = 0;
+
     // read code line
     bool forward_input = false;
 
@@ -595,8 +600,24 @@ static int read_code(const std::string& code, std::vector<PyTorchNode>& nodes)
             if (nscan != 1)
                 continue;
 
-            fprintf(stderr, "netinput = %s\n", netinput);
+//             fprintf(stderr, "netinput = %s\n", netinput);
             forward_input = false;
+
+            PyTorchNode n;
+            n.op = "Input";
+            n.outputs.push_back(netinput);
+
+            {
+                // assign default unknown name
+                char unknownname[256];
+                sprintf(unknownname, "unknownncnn_%d", internal_unknown);
+
+                n.name = unknownname;
+
+                internal_unknown++;
+            }
+
+            nodes.push_back(n);
         }
         else if (strstr(line, " = torch."))
         {
@@ -627,6 +648,17 @@ static int read_code(const std::string& code, std::vector<PyTorchNode>& nodes)
             n.op = op;
             n.outputs = parse_op_output_list(outputs);
             n.args = parse_op_arg_list(op_args.c_str());
+
+            {
+                // assign default unknown name
+                char unknownname[256];
+                sprintf(unknownname, "unknownncnn_%d", internal_unknown);
+
+                n.name = unknownname;
+
+                internal_unknown++;
+            }
+
             nodes.push_back(n);
         }
         else if (strstr(line, "return torch."))
@@ -653,8 +685,19 @@ static int read_code(const std::string& code, std::vector<PyTorchNode>& nodes)
 
             PyTorchNode n;
             n.op = op;
-            n.outputs = parse_op_output_list("ncnnoutput_0");
+            n.outputs.push_back("ncnnoutput_0");
             n.args = parse_op_arg_list(op_args.c_str());
+
+            {
+                // assign default unknown name
+                char unknownname[256];
+                sprintf(unknownname, "unknownncnn_%d", internal_unknown);
+
+                n.name = unknownname;
+
+                internal_unknown++;
+            }
+
             nodes.push_back(n);
         }
     }
@@ -665,6 +708,8 @@ static int read_code(const std::string& code, std::vector<PyTorchNode>& nodes)
 int main(int argc, char** argv)
 {
     const char* ptpath = "model_base.pt";
+    const char* ncnn_prototxt = "ncnn.param";
+    const char* ncnn_modelbin = "ncnn.bin";
 
     std::string code;
     std::string model_json;
@@ -682,26 +727,153 @@ int main(int argc, char** argv)
     std::vector<PyTorchNode> nodes;
     read_code(code, nodes);
 
+    FILE* pp = fopen(ncnn_prototxt, "wb");
+    FILE* bp = fopen(ncnn_modelbin, "wb");
+
+    // magic
+    fprintf(pp, "7767517\n");
+
     const int node_count = nodes.size();
 
     fprintf(stderr, "node_count = %d\n", node_count);
 
+    // node reference
+    std::map<std::string, int> node_reference;
+
+    // weight node
+    std::vector<int> weight_nodes;
+
+    // global definition line
+    // [layer count] [blob count]
+    std::set<std::string> blob_names;
+    for (int i=0; i<node_count; i++)
+    {
+        PyTorchNode& n = nodes[i];
+
+        for (int j=0; j<(int)n.outputs.size(); j++)
+        {
+            blob_names.insert(n.outputs[j]);
+        }
+
+        // distinguish weights and inputs
+        std::vector<std::string> op_arg_list = n.args;
+        std::vector<std::string> input_list;
+        std::vector<std::string> arg_list;
+        for (int i=0; i<(int)op_arg_list.size(); i++)
+        {
+            const std::string& arg = op_arg_list[i];
+            if (blob_names.find(arg) == blob_names.end())
+            {
+                arg_list.push_back(arg);
+            }
+            else
+            {
+                input_list.push_back(arg);
+
+                if (node_reference.find(arg) == node_reference.end())
+                {
+                    node_reference[arg] = 1;
+                }
+                else
+                {
+                    node_reference[arg] = node_reference[arg] + 1;
+                }
+            }
+        }
+
+        n.inputs = input_list;
+        n.args = arg_list;
+    }
+
+    // remove node_reference entry with reference equals to one
+    int splitncnn_blob_count = 0;
+    std::map<std::string, int>::iterator it = node_reference.begin();
+    while (it != node_reference.end())
+    {
+        if (it->second == 1)
+        {
+            node_reference.erase(it++);
+        }
+        else
+        {
+            splitncnn_blob_count += it->second;
+//             fprintf(stderr, "%s %d\n", it->first.c_str(), it->second);
+            ++it;
+        }
+    }
+
+    fprintf(pp, "%lu %lu\n", node_count + node_reference.size(), blob_names.size() + splitncnn_blob_count);
+
+    int internal_split = 0;
+
     for (int i=0; i<node_count; i++)
     {
         const PyTorchNode& n = nodes[i];
 
-        fprintf(stderr, "op = %s\n", n.op.c_str());
+        fprintf(pp, "%-24s", n.op.c_str());
+
+        fprintf(pp, " %-24s %d %d", n.name.c_str(), (int)n.inputs.size(), (int)n.outputs.size());
+
+        for (int j=0; j<(int)n.inputs.size(); j++)
+        {
+            std::string input_name = n.inputs[j];
+
+            if (node_reference.find(input_name) != node_reference.end())
+            {
+                int refidx = node_reference[input_name] - 1;
+                node_reference[input_name] = refidx;
+
+                char splitsuffix[256];
+                sprintf(splitsuffix, "_splitncnn_%d", refidx);
+                input_name = input_name + splitsuffix;
+            }
+
+            fprintf(pp, " %s", input_name.c_str());
+        }
 
         for (int j=0; j<(int)n.outputs.size(); j++)
         {
-            fprintf(stderr, "output = %s\n", n.outputs[j].c_str());
+            fprintf(pp, " %s", n.outputs[j].c_str());
+        }
+
+        // TODO op specific params
+        {
+            for (int j=0; j<(int)n.args.size(); j++)
+            {
+                fprintf(pp, " %s", n.args[j].c_str());
+            }
         }
 
-        for (int j=0; j<(int)n.args.size(); j++)
+        fprintf(pp, "\n");
+
+        for (int j=0; j<(int)n.outputs.size(); j++)
         {
-            fprintf(stderr, "arg = %s\n", n.args[j].c_str());
+            const std::string& output_name = n.outputs[j];
+            if (node_reference.find(output_name) != node_reference.end())
+            {
+                int refcount = node_reference[output_name];
+                if (refcount > 1)
+                {
+                    char splitname[256];
+                    sprintf(splitname, "splitncnn_%d", internal_split);
+                    fprintf(pp, "%-24s %-24s %d %d", "Split", splitname, 1, refcount);
+
+                    fprintf(pp, " %s", output_name.c_str());
+
+                    for (int k=0; k<refcount; k++)
+                    {
+                        fprintf(pp, " %s_splitncnn_%d", output_name.c_str(), k);
+                    }
+                    fprintf(pp, "\n");
+
+                    internal_split++;
+                }
+            }
         }
     }
 
+    fclose(pp);
+    fclose(bp);
+
     return 0;
 }