path: root/nerv/examples/swb_baseline.lua

require 'htk_io'
gconf = {lrate = 0.8, wcost = 1e-6, momentum = 0.9,
        rearrange = true, -- just to make the context order consistent with old results, deprecated
        frm_ext = 5,
        frm_trim = 5, -- trim the first and last 5 frames, TNet just does this, deprecated
        tr_scp = "/slfs1/users/mfy43/swb_ivec/train_bp.scp",
        cv_scp = "/slfs1/users/mfy43/swb_ivec/train_cv.scp",
        htk_conf = "/slfs1/users/mfy43/swb_ivec/plp_0_d_a.conf",
        initialized_param = {"/slfs1/users/mfy43/swb_init.nerv",
                            "/slfs1/users/mfy43/swb_global_transf.nerv"}}

function make_layer_repo(param_repo)
    local layer_repo = nerv.LayerRepo(
    {
        -- global transf
        ["nerv.BiasLayer"] =
        {
            blayer1 = {dim_in = {429}, dim_out = {429}, params = {bias = "bias1"}},
            blayer2 = {dim_in = {429}, dim_out = {429}, params = {bias = "bias2"}}
        },
        ["nerv.WindowLayer"] =
        {
            wlayer1 = {dim_in = {429}, dim_out = {429}, params = {window = "window1"}},
            wlayer2 = {dim_in = {429}, dim_out = {429}, params = {window = "window2"}}
        },
        -- biased linearity
        ["nerv.AffineLayer"] =
        {
            affine0 = {dim_in = {429}, dim_out = {2048},
                        params = {ltp = "affine0_ltp", bp = "affine0_bp"}},
            affine1 = {dim_in = {2048}, dim_out = {2048},
                        params = {ltp = "affine1_ltp", bp = "affine1_bp"}},
            affine2 = {dim_in = {2048}, dim_out = {2048},
                        params = {ltp = "affine2_ltp", bp = "affine2_bp"}},
            affine3 = {dim_in = {2048}, dim_out = {2048},
                        params = {ltp = "affine3_ltp", bp = "affine3_bp"}},
            affine4 = {dim_in = {2048}, dim_out = {2048},
                        params = {ltp = "affine4_ltp", bp = "affine4_bp"}},
            affine5 = {dim_in = {2048}, dim_out = {2048},
                        params = {ltp = "affine5_ltp", bp = "affine5_bp"}},
            affine6 = {dim_in = {2048}, dim_out = {2048},
                        params = {ltp = "affine6_ltp", bp = "affine6_bp"}},
            affine7 = {dim_in = {2048}, dim_out = {3001},
                        params = {ltp = "affine7_ltp", bp = "affine7_bp"}}
        },
        ["nerv.SigmoidLayer"] =
        {
            sigmoid0 = {dim_in = {2048}, dim_out = {2048}},
            sigmoid1 = {dim_in = {2048}, dim_out = {2048}},
            sigmoid2 = {dim_in = {2048}, dim_out = {2048}},
            sigmoid3 = {dim_in = {2048}, dim_out = {2048}},
            sigmoid4 = {dim_in = {2048}, dim_out = {2048}},
            sigmoid5 = {dim_in = {2048}, dim_out = {2048}},
            sigmoid6 = {dim_in = {2048}, dim_out = {2048}}
        },
        ["nerv.SoftmaxCELayer"] = -- softmax + ce criterion layer for finetune output
        {
            ce_crit = {dim_in = {3001, 1}, dim_out = {1}, compressed = true}
        },
        ["nerv.SoftmaxLayer"] = -- softmax for decode output
        {
            softmax = {dim_in = {3001}, dim_out = {3001}}
        }
    }, param_repo, gconf)

    layer_repo:add_layers(
    {
        ["nerv.DAGLayer"] =
        {
            global_transf = {
                dim_in