start implementing tdag_layer

2 files changed, 455 insertions, 0 deletions

diff --git a/nerv/examples/lmptb/m-tests/dagl_test.lua b/nerv/examples/lmptb/m-tests/dagl_test.lua
new file mode 100644
index 0000000..02e9c49
--- /dev/null
+++ b/nerv/examples/lmptb/m-tests/dagl_test.lua
@@ -0,0 +1,164 @@
+require 'lmptb.lmvocab'
+require 'lmptb.lmfeeder'
+require 'lmptb.lmutil'
+require 'lmptb.layer.init'
+require 'rnn.layer_tdag'
+
+--[[global function rename]]--
+printf = nerv.printf
+--[[global function rename ends]]--
+
+--global_conf: table
+--first_time: bool
+--Returns: a ParamRepo
+function prepare_parameters(global_conf, first_time)
+    printf("%s preparing parameters...\n", global_conf.sche_log_pre) 
+    
+    if (first_time) then
+        ltp_ih = nerv.LinearTransParam("ltp_ih", global_conf)  
+        ltp_ih.trans = global_conf.cumat_type(global_conf.vocab:size() + 1, global_conf.hidden_size) --index 0 is for zero, others correspond to vocab index(starting from 1)
+        ltp_ih.trans:generate(global_conf.param_random)
+        ltp_ih.trans[0]:fill(0)
+
+        ltp_hh = nerv.LinearTransParam("ltp_hh", global_conf)
+        ltp_hh.trans = global_conf.cumat_type(global_conf.hidden_size, global_conf.hidden_size)
+        ltp_hh.trans:generate(global_conf.param_random) 
+
+        ltp_ho = nerv.LinearTransParam("ltp_ho", global_conf)
+        ltp_ho.trans = global_conf.cumat_type(global_conf.hidden_size, global_conf.vocab:size())
+        ltp_ho.trans:generate(global_conf.param_random)
+
+        bp_h = nerv.BiasParam("bp_h", global_conf)
+        bp_h.trans = global_conf.cumat_type(1, global_conf.hidden_size)
+        bp_h.trans:generate(global_conf.param_random)
+
+        bp_o = nerv.BiasParam("bp_o", global_conf)
+        bp_o.trans = global_conf.cumat_type(1, global_conf.vocab:size())
+        bp_o.trans:generate(global_conf.param_random)
+
+        local f = nerv.ChunkFile(global_conf.param_fn, 'w')
+        f:write_chunk(ltp_ih)
+        f:write_chunk(ltp_hh)
+        f:write_chunk(ltp_ho)
+        f:write_chunk(bp_h)
+        f:write_chunk(bp_o)
+        f:close()
+    end
+    
+    local paramRepo = nerv.ParamRepo()
+    paramRepo:import({global_conf.param_fn}, nil, global_conf)
+
+    printf("%s preparing parameters end.\n", global_conf.sche_log_pre)
+
+    return paramRepo
+end
+
+--global_conf: table
+--Returns: nerv.LayerRepo
+function prepare_layers(global_conf, paramRepo)
+    printf("%s preparing layers...\n", global_conf.sche_log_pre)
+
+    local recurrentLconfig = {{["bp"] = "bp_h", ["ltp_hh"] = "ltp_hh"}, {["dim_in"] = {global_conf.hidden_size, global_conf.hidden_size}, ["dim_out"] = {global_conf.hidden_size}, ["break_id"] = global_conf.vocab:get_sen_entry().id, ["independent"] = global_conf.independent, ["clip"] = 10}}
+
+    local layers = {
+        ["nerv.IndRecurrentLayer"] = {
+            ["recurrentL1"] = recurrentLconfig, 
+        },
+
+        ["nerv.SelectLinearLayer"] = {
+            ["selectL1"] = {{["ltp"] = "ltp_ih"}, {["dim_in"] = {1}, ["dim_out"] = {global_conf.hidden_size}}},
+        },
+
+        ["nerv.SigmoidLayer"] = {
+            ["sigmoidL1"] = {{}, {["dim_in"] = {global_conf.hidden_size}, ["dim_out"] = {global_conf.hidden_size}}}
+        },
+        
+        ["nerv.AffineLayer"] = {
+            ["outputL"] = {{["ltp"] = "ltp_ho", ["bp"] = "bp_o"}, {["dim_in"] = {global_conf.hidden_size}, ["dim_out"] = {global_conf.vocab:size()}}},
+        },
+
+        ["nerv.SoftmaxCELayer"] = {
+            ["softmaxL"] = {{}, {["dim_in"] = {global_conf.vocab:size(), global_conf.vocab:size()}, ["dim_out"] = {1}}},
+        },
+    }
+    
+    --[[ --we do not need those in the new rnn framework
+    printf("%s adding %d bptt layers...\n", global_conf.sche_log_pre, global_conf.bptt)
+    for i = 1, global_conf.bptt do
+        layers["nerv.IndRecurrentLayer"]["recurrentL" .. (i + 1)] = recurrentLconfig 
+        layers["nerv.SigmoidLayer"]["sigmoidL" .. (i + 1)] = {{}, {["dim_in"] = {global_conf.hidden_size}, ["dim_out"] = {global_conf.hidden_size}}}
+        layers["nerv.SelectLinearLayer"]["selectL" .. (i + 1)] = {{["ltp"] = "ltp_ih"}, {["dim_in"] = {1}, ["dim_out"] = {global_conf.hidden_size}}}
+    end
+    --]]
+
+    local layerRepo = nerv.LayerRepo(layers, paramRepo, global_conf)
+    printf("%s preparing layers end.\n", global_conf.sche_log_pre)
+    return layerRepo
+end
+
+--global_conf: table
+--layerRepo: nerv.LayerRepo
+--Returns: a nerv.DAGLayer
+function prepare_dagLayer(global_conf, layerRepo)
+    printf("%s Initing daglayer ...\n", global_conf.sche_log_pre)
+
+    --input: input_w, input_w, ... input_w_now, last_activation
+    local dim_in_t = {}
+    dim_in_t[1] = 1 --input to select_linear layer
+    dim_in_t[2] = global_conf.vocab:size() --input to softmax label
+    local connections_t = {
+                ["<input>[1]"] = "selectL1[1],0",
+                ["selectL1[1]"] = "recurrentL1[1],0", 
+                ["recurrentL1[1]"] = "sigmoidL1[1],0",
+                ["sigmoidL1[1]"] = "outputL[1],0",
+                ["sigmoidL1[1]"] = "recurrentL1[2],1",
+                ["outputL[1]"] = "softmaxL[1],0",
+                ["<input>[2]"] = "softmaxL[2],0",
+                ["softmaxL[1]"] = "<output>[1],0"
+    }
+
+    --[[
+    printf("%s printing DAG connections:\n", global_conf.sche_log_pre)
+    for key, value in pairs(connections_t) do
+        printf("\t%s->%s\n", key, value)
+    end
+    ]]--
+
+    local dagL = nerv.TDAGLayer("dagL", global_conf, {["dim_in"] = dim_in_t, ["dim_out"] = {1}, ["sub_layers"] = layerRepo,
+            ["connections"] = connections_t, 
+        })
+    dagL:init(global_conf.batch_size)
+    printf("%s Initing DAGLayer end.\n", global_conf.sche_log_pre)
+    return dagL
+end
+
+train_fn = '/home/slhome/txh18/workspace/nerv/nerv/nerv/examples/lmptb/m-tests/some-text'
+test_fn = '/home/slhome/txh18/workspace/nerv/nerv/nerv/examples/lmptb/m-tests/some-text'
+
+global_conf = {
+    lrate = 1, wcost = 1e-6, momentum = 0,
+    cumat_type = nerv.CuMatrixFloat,
+    mmat_type = nerv.CuMatrixFloat,
+
+    hidden_size = 20,
+    batch_size = 5,
+    seq_size = 3, 
+    max_iter = 18,
+    param_random = function() return (math.random() / 5 - 0.1) end,
+    independent = true,
+
+    train_fn = train_fn,
+    test_fn = test_fn,
+    sche_log_pre = "[SCHEDULER]:",
+    log_w_num = 10, --give a message when log_w_num words have been processed
+    timer = nerv.Timer()
+}
+global_conf.work_dir = '/home/slhome/txh18/workspace/nerv/play/dagL_test'
+global_conf.param_fn = global_conf.work_dir.."/params"
+
+local vocab = nerv.LMVocab()
+global_conf["vocab"] = vocab
+global_conf.vocab:build_file(global_conf.train_fn, false)
+local paramRepo = prepare_parameters(global_conf, true)
+local layerRepo = prepare_layers(global_conf, paramRepo)
+local dagL = prepare_dagLayer(global_conf, layerRepo)
diff --git a/nerv/examples/lmptb/rnn/layer_tdag.lua b/nerv/examples/lmptb/rnn/layer_tdag.lua
new file mode 100644
index 0000000..296e2e6
--- /dev/null
+++ b/nerv/examples/lmptb/rnn/layer_tdag.lua
@@ -0,0 +1,291 @@
+local DAGLayer = nerv.class("nerv.TDAGLayer", "nerv.Layer")
+
+local function parse_id(str)
+    local id, port, time, _
+    _, _, id, port, time = string.find(str, "([a-zA-Z0-9_]+)%[([0-9]+)%][,]*([0-9]*)")
+    if id == nil or port == nil then
+        _, _, id, port, time = string.find(str, "(.+)%[([0-9]+)%][,]*([0-9]*)")
+        if not (id == "<input>" or id == "<output>") then
+            nerv.error("wrong format of connection id")
+        end
+    end
+    print(str, id, port, time)
+    port = tonumber(port)
+    if (time == nil) then
+        time = 0
+    else
+        time = tonumber(time)
+    end
+    return id, port, time
+end
+
+local function discover(id, layers, layer_repo)
+    local ref = layers[id]
+    if id == "<input>" or id == "<output>" then
+        return nil
+    end
+    if ref == nil then
+        local layer = layer_repo:get_layer(id)
+        local dim_in, dim_out = layer:get_dim()
+        ref = {
+            layer = layer,
+            inputs = {},
+            outputs = {},
+            err_inputs = {},
+            err_outputs = {},
+            next_layers = {},
+            input_len = #dim_in,
+            output_len = #dim_out,
+            in_deg = 0,
+            visited = false
+        }
+        layers[id] = ref
+    end
+    return ref
+end
+
+function DAGLayer:__init(id, global_conf, layer_conf)
+    local layers = {}
+    local inputs = {}
+    local outputs = {}
+    local dim_in = layer_conf.dim_in
+    local dim_out = layer_conf.dim_out
+    local parsed_conn = {}
+    local _
+    local time_to
+
+    for from, to in pairs(layer_conf.connections) do
+        
+        local id_from, port_from, _ = parse_id(from)
+        local id_to, port_to, time_to = parse_id(to)
+        
+        local ref_from = discover(id_from, layers, layer_conf.sub_layers)
+        local ref_to = discover(id_to, layers, layer_conf.sub_layers)
+        
+        local input_dim, output_dim, _
+        
+        if ref_from and ref_from.outputs[port_from] ~= nil then
+            nerv.error("%s has already been attached", from)
+        end
+        if ref_to and ref_to.inputs[port_to] ~= nil then
+            nerv.error("%s has already been attached", to)
+        end
+        if id_from == "<input>" then
+            input_dim, _ = ref_to.layer:get_dim()
+            if dim_in[port_from] ~= input_dim[port_to] then
+                nerv.error("mismatching data dimension between %s and %s", from, to)
+            end
+            inputs[port_from] = {ref_to, port_to}
+            ref_to.inputs[port_to] = inputs -- just a place holder
+        elseif id_to == "<output>" then
+            _, output_dim = ref_from.layer:get_dim()
+            if output_dim[port_from] ~= dim_out[port_to] then
+                nerv.error("mismatching data dimension between %s and %s", from, to)
+            end
+            outputs[port_to] = {ref_from, port_from}
+            ref_from.outputs[port_from] = outputs -- just a place holder
+        else
+            _, output_dim = ref_from.layer:get_dim()
+            input_dim, _ = ref_to.layer:get_dim()
+            if output_dim[port_from] ~= input_dim[port_to] then
+                nerv.error("mismatching data dimension between %s and %s", from, to)
+            end
+
+            table.insert(parsed_conn,
+            {{ref_from, port_from}, {ref_to, port_to}})
+            table.insert(ref_from.next_layers, ref_to) -- add edge
+            ref_to.in_deg = ref_to.in_deg + 1          -- increase the in-degree of the target layer
+        end
+    end
+
+    self.layers = layers
+    self.inputs = inputs
+    self.outputs = outputs
+    self.id = id
+    self.dim_in = dim_in
+    self.dim_out = dim_out
+    self.parsed_conn = parsed_conn
+    self.queue = queue
+    self.gconf = global_conf
+end
+
+function DAGLayer:init(batch_size)
+    for i, conn in ipairs(self.parsed_conn) do
+        local _, output_dim
+        local ref_from, port_from, ref_to, port_to
+        ref_from, port_from = unpack(conn[1])
+        ref_to, port_to = unpack(conn[2])
+        _, output_dim = ref_from.layer:get_dim()
+        local dim = 1
+        if output_dim[port_from] > 0 then
+            dim = output_dim[port_from]
+        end
+        local mid = self.gconf.cumat_type(batch_size, dim)
+        local err_mid = mid:create()
+
+        ref_from.outputs[port_from] = mid
+        ref_to.inputs[port_to] = mid
+
+        ref_from.err_inputs[port_from] = err_mid
+        ref_to.err_outputs[port_to] = err_mid
+    end
+    for id, ref in pairs(self.layers) do
+        for i = 1, ref.input_len do
+            if ref.inputs[i] == nil then
+                nerv.error("dangling input port %d of layer %s", i, id)
+            end
+        end
+        for i = 1, ref.output_len do
+            if ref.outputs[i] == nil then
+                nerv.error("dangling output port %d of layer %s", i, id)
+            end
+        end
+        -- initialize sub layers
+        ref.layer:init(batch_size)
+    end
+    for i = 1, #self.dim_in do
+        if self.inputs[i] == nil then
+            nerv.error("dangling port %d of layer <input>", i)
+        end
+    end
+    for i = 1, #self.dim_out do
+        if self.outputs[i] == nil then
+            nerv.error("dangling port %d of layer <output>", i)
+        end
+    end
+end
+
+function DAGLayer:batch_resize(batch_size)
+    self.gconf.batch_size = batch_size
+
+    for i, conn in ipairs(self.parsed_conn) do
+        local _, output_dim
+        local ref_from, port_from, ref_to, port_to
+        ref_from, port_from = unpack(conn[1])
+        ref_to, port_to = unpack(conn[2])
+        _, output_dim = ref_from.layer:get_dim()
+
+        if ref_from.outputs[port_from]:nrow() ~= batch_size and output_dim[port_from] > 0 then
+            local mid = self.gconf.cumat_type(batch_size, output_dim[port_from])
+            local err_mid = mid:create()
+
+            ref_from.outputs[port_from] = mid
+            ref_to.inputs[port_to] = mid
+
+            ref_from.err_inputs[port_from] = err_mid
+            ref_to.err_outputs[port_to] = err_mid
+        end
+    end
+    for id, ref in pairs(self.layers) do
+        ref.layer:batch_resize(batch_size)
+    end
+    collectgarbage("collect")
+end
+
+function DAGLayer:set_inputs(input)
+    for i = 1, #self.dim_in do
+        if input[i] == nil then
+            nerv.error("some input is not provided");
+        end
+        local layer = self.inputs[i][1]
+        local port = self.inputs[i][2]
+        layer.inputs[port] = input[i]
+    end
+end
+
+function DAGLayer:set_outputs(output)
+    for i = 1, #self.dim_out do
+        if output[i] == nil then
+            nerv.error("some output is not provided");
+        end
+        local layer = self.outputs[i][1]
+        local port = self.outputs[i][2]
+        layer.outputs[port] = output[i]
+    end
+end
+
+function DAGLayer:set_err_inputs(bp_err)
+    for i = 1, #self.dim_out do
+        local layer = self.outputs[i][1]
+        local port = self.outputs[i][2]
+        layer.err_inputs[port] = bp_err[i]
+    end
+end
+
+function DAGLayer:set_err_outputs(next_bp_err)
+    for i = 1, #self.dim_in do
+        local layer = self.inputs[i][1]
+        local port = self.inputs[i][2]
+        layer.err_outputs[port] = next_bp_err[i]
+    end
+end
+
+function DAGLayer:update(bp_err, input, output)
+    self:set_err_inputs(bp_err)
+    self:set_inputs(input)
+    self:set_outputs(output)
+    -- print("update")
+    for id, ref in pairs(self.queue) do
+        -- print(ref.layer.id)
+        ref.layer:update(ref.err_inputs, ref.inputs, ref.outputs)
+    end
+end
+
+function DAGLayer:propagate(input, output)
+    self:set_inputs(input)
+    self:set_outputs(output)
+    local ret = false
+    for i = 1, #self.queue do
+        local ref = self.queue[i]
+        -- print(ref.layer.id)
+        ret = ref.layer:propagate(ref.inputs, ref.outputs)
+    end
+    return ret
+end
+
+function DAGLayer:back_propagate(bp_err, next_bp_err, input, output)
+    self:set_err_outputs(next_bp_err)
+    self:set_err_inputs(bp_err)
+    self:set_inputs(input)
+    self:set_outputs(output)
+    for i = #self.queue, 1, -1 do
+        local ref = self.queue[i]
+        -- print(ref.layer.id)
+        ref.layer:back_propagate(ref.err_inputs, ref.err_outputs, ref.inputs, ref.outputs)
+    end
+end
+
+function DAGLayer:get_params()
+    local param_repos = {}
+    for id, ref in pairs(self.queue) do
+        table.insert(param_repos, ref.layer:get_params())
+    end
+    return nerv.ParamRepo.merge(param_repos)
+end
+
+DAGLayer.PORT_TYPES = {
+    INPUT = {},
+    OUTPUT = {},
+    ERR_INPUT = {},
+    ERR_OUTPUT = {}
+}
+
+function DAGLayer:get_intermediate(id, port_type)
+    if id == "<input>" or id == "<output>" then
+        nerv.error("an actual real layer id is expected")
+    end
+    local layer = self.layers[id]
+    if layer == nil then
+        nerv.error("layer id %s not found", id)
+    end
+    if port_type == DAGLayer.PORT_TYPES.INPUT then
+        return layer.inputs
+    elseif port_type == DAGLayer.PORT_TYPES.OUTPUT then
+        return layer.outputs
+    elseif port_type == DAGLayer.PORT_TYPES.ERR_INPUT then
+        return layer.err_inputs
+    elseif port_type == DAGLayer.PORT_TYPES.ERR_OUTPUT then
+        return layer.err_outputs
+    end
+    nerv.error("unrecognized port type")
+end