implementing GateFFF layer

1 files changed, 33 insertions, 32 deletions

diff --git a/nerv/examples/lmptb/rnn/layers/gate_fff.lua b/nerv/examples/lmptb/rnn/layers/gate_fff.lua
index 6a588fc..1010639 100644
--- a/nerv/examples/lmptb/rnn/layers/gate_fff.lua
+++ b/nerv/examples/lmptb/rnn/layers/gate_fff.lua
@@ -2,64 +2,65 @@ local GateFFFLayer = nerv.class('nerv.GateFFFLayer', 'nerv.Layer')
 
 function GateFFFLayer:__init(id, global_conf, layer_conf)
     self.id = id
-    self.ltp = layer_conf.ltp
-    self.bp = layer_conf.bp
     self.dim_in = layer_conf.dim_in
     self.dim_out = layer_conf.dim_out
     self.gconf = global_conf
-    self:check_dim_len(1, 1) -- exactly one input and one output
+ 
+    self.ltp1 = self:find_param("ltp1", layer_conf, global_conf, nerv.LinearTransParam, {self.dim_in[1], self.dim_out[1]}) --layer_conf.ltp
+    self.ltp2 = self:find_param("ltp2", layer_conf, global_conf, nerv.LinearTransParam, {self.dim_in[2], self.dim_out[1]}) --layer_conf.ltp
+    self.ltp3 = self:find_param("ltp3", layer_conf, global_conf, nerv.LinearTransParam, {self.dim_in[3], self.dim_out[1]}) --layer_conf.ltp
+    self.bp = self:find_param("bp", layer_conf, global_conf, nerv.BiasParam, {1, self.dim_out[1]})--layer_conf.bp
+  
+    self:check_dim_len(3, 1) -- exactly one input and one output
 end
 
 function GateFFFLayer:init(batch_size)
-    if self.ltp.trans:ncol() ~= self.bp.trans:ncol() then
+    if self.ltp1.trans:ncol() ~= self.bp.trans:ncol() or 
+        self.ltp2.trans:ncol() ~= self.bp.trans:ncol() or
+        self.ltp3.trans:ncol() ~= self.bp.trans:ncol() then
         nerv.error("mismatching dimensions of linear transform and bias paramter")
     end
-    if self.dim_in[1] ~= self.ltp.trans:nrow() then
+    if self.dim_in[1] ~= self.ltp1.trans:nrow() or
+        self.dim_in[2] ~= self.ltp2.trans:nrow() or
+        self.dim_in[3] ~= self.ltp3.trans:nrow() then
         nerv.error("mismatching dimensions of linear transform parameter and input")
     end
-    if self.dim_out[1] ~= self.ltp.trans:ncol() then
+    if self.dim_out[1] ~= self.ltp1.trans:ncol() then
         nerv.error("mismatching dimensions of linear transform parameter and output")
     end
-    self.ltp_grad = self.ltp.trans:create()
-    self.ltp:train_init()
+    self.ltp1:train_init()
+    self.ltp2:train_init()
+    self.ltp3:train_init()
     self.bp:train_init()
+    self.err_bakm = self.gconf.cumat_type(batch_size, self.dim_out[1])
 end
 
 function GateFFFLayer:batch_resize(batch_size)
-    -- do nothing
-end
-
-function GateFFFLayer:update(bp_err, input, output)
-    if self.direct_update == true then
-        local gconf = self.gconf
-        if gconf.momentum > 0 then
-            self.ltp.correction:mul(input[1], bp_err[1], 1.0, gconf.momentum, 'T', 'N')
-            self.bp.correction:add(self.bp.correction, bp_err[1]:colsum(), gconf.momentum, 1)
-            -- momentum gain
-            local mmt_gain = 1.0 / (1.0 - gconf.momentum)
-            local n = self.gconf.batch_size * mmt_gain
-            -- perform update
-            self.ltp.trans:add(self.ltp.trans, self.ltp.correction, 1.0 - gconf.lrate * gconf.wcost / gconf.batch_size, - gconf.lrate / n)
-            self.bp.trans:add(self.bp.trans, self.bp.correction, 1.0 - gconf.lrate * gconf.wcost / gconf.batch_size, - gconf.lrate / n)
-        else
-            self.ltp.trans:mul(input[1], bp_err[1], - gconf.lrate / gconf.batch_size, 1.0 - gconf.lrate * gconf.wcost / gconf.batch_size, 'T', 'N')
-            self.bp.trans:add(self.bp.trans, bp_err[1]:colsum(), 1.0 - gconf.lrate * gconf.wcost / gconf.batch_size, - gconf.lrate / gconf.batch_size)
-        end
-    else
-        self.ltp:update_by_err_input(bp_err[1], input[1])
-        self.bp:update_by_gradient(bp_err[1]:colsum())
+    if self.err_m:nrow() ~= batch_size then
+        self.err_bakm = self.gconf.cumat_type(batch_size, self.dim_out[1])
     end
 end
 
 function GateFFFLayer:propagate(input, output)
     -- apply linear transform
-    output[1]:mul(input[1], self.ltp.trans, 1.0, 0.0, 'N', 'N')
+    output[1]:mul(input[1], self.ltp1.trans, 1.0, 0.0, 'N', 'N')
+    output[1]:mul(input[2], self.ltp2.trans, 1.0, 1.0, 'N', 'N')
+    output[1]:mul(input[3], self.ltp3.trans, 1.0, 1.0, 'N', 'N')
     -- add bias
     output[1]:add_row(self.bp.trans, 1.0)
+    output[1]:sigmoid(output[1])
 end
 
 function GateFFFLayer:back_propagate(bp_err, next_bp_err, input, output)
-    next_bp_err[1]:mul(bp_err[1], self.ltp.trans, 1.0, 0.0, 'N', 'T')
+    self.err_bakm:sigmoid_grad(bp_err[1], output[1])
+    next_bp_err[1]:mul(self.err_bakm, self.ltp1.trans, 1.0, 0.0, 'N', 'T')
+    next_bp_err[2]:mul(self.err_bakm, self.ltp2.trans, 1.0, 0.0, 'N', 'T')
+    next_bp_err[3]:mul(self.err_bakm, self.ltp3.trans, 1.0, 0.0, 'N', 'T')
+end
+
+function GateFFFLayer:update(bp_err, input, output)
+    self.ltp:update_by_err_input(bp_err[1], input[1])
+    self.bp:update_by_gradient(bp_err[1]:colsum())
 end
 
 function GateFFFLayer:get_params()