eden-emu/opus
19
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

DRED: quantize scale and dead zone to 8 bits

Browse source
This commit is contained in:
Jean-Marc Valin 2023-11-07 17:46:38 -05:00
parent 4e104555e9
commit 222662dac8
No known key found for this signature in database
GPG key ID: 531A52533318F00A
4 changed files with 23 additions and 17 deletions
Download patch file Download diff file Expand all files Collapse all files

2
autogen.sh
View file

@ -9,7 +9,7 @@ set -e
srcdir=`dirname $0` srcdir=`dirname $0`
test -n "$srcdir" && cd "$srcdir" test -n "$srcdir" && cd "$srcdir"
dnn/download_model.sh 2386a60 dnn/download_model.sh b6095cf
echo "Updating build configuration files, please wait...." echo "Updating build configuration files, please wait...."

28
dnn/torch/rdovae/export_rdovae_weights.py
View file

@ -59,33 +59,35 @@ def dump_statistical_model(writer, w, name):
p0 = torch.sigmoid(w[:, 4 , :]).numpy() p0 = torch.sigmoid(w[:, 4 , :]).numpy()
p0 = 1 - r ** (0.5 + 0.5 * p0) p0 = 1 - r ** (0.5 + 0.5 * p0)
scales_norm = 255./256./(1e-15+np.max(quant_scales,axis=0))
quant_scales = quant_scales*scales_norm
quant_scales_q8 = np.round(quant_scales * 2**8).astype(np.uint16) quant_scales_q8 = np.round(quant_scales * 2**8).astype(np.uint16)
dead_zone_q10 = np.round(dead_zone * 2**10).astype(np.uint16) dead_zone_q8 = np.clip(np.round(dead_zone * 2**8), 0, 255).astype(np.uint16)
r_q8 = np.clip(np.round(r * 2**8), 0, 255).astype(np.uint8) r_q8 = np.clip(np.round(r * 2**8), 0, 255).astype(np.uint8)
p0_q8 = np.clip(np.round(p0 * 2**8), 0, 255).astype(np.uint16) p0_q8 = np.clip(np.round(p0 * 2**8), 0, 255).astype(np.uint16)
mask = (np.max(r_q8,axis=0) > 0) * (np.min(p0_q8,axis=0) < 255) mask = (np.max(r_q8,axis=0) > 0) * (np.min(p0_q8,axis=0) < 255)
quant_scales_q8 = quant_scales_q8[:, mask] quant_scales_q8 = quant_scales_q8[:, mask]
dead_zone_q10 = dead_zone_q10[:, mask] dead_zone_q8 = dead_zone_q8[:, mask]
r_q8 = r_q8[:, mask] r_q8 = r_q8[:, mask]
p0_q8 = p0_q8[:, mask] p0_q8 = p0_q8[:, mask]
N = r_q8.shape[-1] N = r_q8.shape[-1]
print_vector(writer.source, quant_scales_q8, f'dred_{name}_quant_scales_q8', dtype='opus_uint16', static=False) print_vector(writer.source, quant_scales_q8, f'dred_{name}_quant_scales_q8', dtype='opus_uint8', static=False)
print_vector(writer.source, dead_zone_q10, f'dred_{name}_dead_zone_q10', dtype='opus_uint16', static=False) print_vector(writer.source, dead_zone_q8, f'dred_{name}_dead_zone_q8', dtype='opus_uint8', static=False)
print_vector(writer.source, r_q8, f'dred_{name}_r_q8', dtype='opus_uint8', static=False) print_vector(writer.source, r_q8, f'dred_{name}_r_q8', dtype='opus_uint8', static=False)
print_vector(writer.source, p0_q8, f'dred_{name}_p0_q8', dtype='opus_uint8', static=False) print_vector(writer.source, p0_q8, f'dred_{name}_p0_q8', dtype='opus_uint8', static=False)
writer.header.write( writer.header.write(
f""" f"""
extern const opus_uint16 dred_{name}_quant_scales_q8[{levels * N}]; extern const opus_uint8 dred_{name}_quant_scales_q8[{levels * N}];
extern const opus_uint16 dred_{name}_dead_zone_q10[{levels * N}]; extern const opus_uint8 dred_{name}_dead_zone_q8[{levels * N}];
extern const opus_uint8 dred_{name}_r_q8[{levels * N}]; extern const opus_uint8 dred_{name}_r_q8[{levels * N}];
extern const opus_uint8 dred_{name}_p0_q8[{levels * N}]; extern const opus_uint8 dred_{name}_p0_q8[{levels * N}];
""" """
) )
return N, mask return N, mask, torch.tensor(scales_norm[mask])
def c_export(args, model): def c_export(args, model):
@ -128,14 +130,16 @@ f"""
levels = qembedding.shape[0] levels = qembedding.shape[0]
qembedding = torch.reshape(qembedding, (levels, 6, -1)) qembedding = torch.reshape(qembedding, (levels, 6, -1))
latent_dim, latent_mask = dump_statistical_model(stats_writer, qembedding[:, :, :orig_latent_dim], 'latent') latent_dim, latent_mask, latent_scale = dump_statistical_model(stats_writer, qembedding[:, :, :orig_latent_dim], 'latent')
state_dim, state_mask = dump_statistical_model(stats_writer, qembedding[:, :, orig_latent_dim:], 'state') state_dim, state_mask, state_scale = dump_statistical_model(stats_writer, qembedding[:, :, orig_latent_dim:], 'state')
padded_latent_dim = (latent_dim+7)//8*8 padded_latent_dim = (latent_dim+7)//8*8
latent_pad = padded_latent_dim - latent_dim; latent_pad = padded_latent_dim - latent_dim;
w = latent_out.weight[latent_mask,:] w = latent_out.weight[latent_mask,:]
w = w/latent_scale[:, None]
w = torch.cat([w, torch.zeros(latent_pad, w.shape[1])], dim=0) w = torch.cat([w, torch.zeros(latent_pad, w.shape[1])], dim=0)
b = latent_out.bias[latent_mask] b = latent_out.bias[latent_mask]
b = b/latent_scale
b = torch.cat([b, torch.zeros(latent_pad)], dim=0) b = torch.cat([b, torch.zeros(latent_pad)], dim=0)
latent_out.weight = torch.nn.Parameter(w) latent_out.weight = torch.nn.Parameter(w)
latent_out.bias = torch.nn.Parameter(b) latent_out.bias = torch.nn.Parameter(b)
@ -143,16 +147,18 @@ f"""
padded_state_dim = (state_dim+7)//8*8 padded_state_dim = (state_dim+7)//8*8
state_pad = padded_state_dim - state_dim; state_pad = padded_state_dim - state_dim;
w = state_out.weight[state_mask,:] w = state_out.weight[state_mask,:]
w = w/state_scale[:, None]
w = torch.cat([w, torch.zeros(state_pad, w.shape[1])], dim=0) w = torch.cat([w, torch.zeros(state_pad, w.shape[1])], dim=0)
b = state_out.bias[state_mask] b = state_out.bias[state_mask]
b = b/state_scale
b = torch.cat([b, torch.zeros(state_pad)], dim=0) b = torch.cat([b, torch.zeros(state_pad)], dim=0)
state_out.weight = torch.nn.Parameter(w) state_out.weight = torch.nn.Parameter(w)
state_out.bias = torch.nn.Parameter(b) state_out.bias = torch.nn.Parameter(b)
latent_in = model.get_submodule('core_decoder.module.dense_1') latent_in = model.get_submodule('core_decoder.module.dense_1')
state_in = model.get_submodule('core_decoder.module.hidden_init') state_in = model.get_submodule('core_decoder.module.hidden_init')
latent_in.weight = torch.nn.Parameter(latent_in.weight[:,latent_mask]) latent_in.weight = torch.nn.Parameter(latent_in.weight[:,latent_mask]*latent_scale)
state_in.weight = torch.nn.Parameter(state_in.weight[:,state_mask]) state_in.weight = torch.nn.Parameter(state_in.weight[:,state_mask]*state_scale)
# encoder # encoder
encoder_dense_layers = [ encoder_dense_layers = [

2
silk/dred_decoder.c
View file

@ -45,7 +45,7 @@ static int sign_extend(int x, int b) {
return (x ^ m) - m; return (x ^ m) - m;
} }
static void dred_decode_latents(ec_dec *dec, float *x, const opus_uint16 *scale, const opus_uint8 *r, const opus_uint8 *p0, int dim) { static void dred_decode_latents(ec_dec *dec, float *x, const opus_uint8 *scale, const opus_uint8 *r, const opus_uint8 *p0, int dim) {
int i; int i;
for (i=0;i<dim;i++) { for (i=0;i<dim;i++) {
int q; int q;

8
silk/dred_encoder.c
View file

@ -223,7 +223,7 @@ void dred_compute_latents(DREDEnc *enc, const float *pcm, int frame_size, int ex
} }
} }
static void dred_encode_latents(ec_enc *enc, const float *x, const opus_uint16 *scale, const opus_uint16 *dzone, const opus_uint8 *r, const opus_uint8 *p0, int dim) { static void dred_encode_latents(ec_enc *enc, const float *x, const opus_uint8 *scale, const opus_uint8 *dzone, const opus_uint8 *r, const opus_uint8 *p0, int dim) {
int i; int i;
int q[IMAX(DRED_LATENT_DIM,DRED_STATE_DIM)]; int q[IMAX(DRED_LATENT_DIM,DRED_STATE_DIM)];
float xq[IMAX(DRED_LATENT_DIM,DRED_STATE_DIM)]; float xq[IMAX(DRED_LATENT_DIM,DRED_STATE_DIM)];
@ -233,7 +233,7 @@ static void dred_encode_latents(ec_enc *enc, const float *x, const opus_uint16 *
/* This is split into multiple loops (with temporary arrays) so that the compiler /* This is split into multiple loops (with temporary arrays) so that the compiler
can vectorize all of it, and so we can call the vector tanh(). */ can vectorize all of it, and so we can call the vector tanh(). */
for (i=0;i<dim;i++) { for (i=0;i<dim;i++) {
delta[i] = dzone[i]*(1.f/1024.f); delta[i] = dzone[i]*(1.f/256.f);
xq[i] = x[i]*scale[i]*(1.f/256.f); xq[i] = x[i]*scale[i]*(1.f/256.f);
deadzone[i] = xq[i]/(delta[i]+eps); deadzone[i] = xq[i]/(delta[i]+eps);
} }
@ -272,7 +272,7 @@ int dred_encode_silk_frame(const DREDEnc *enc, unsigned char *buf, int max_chunk
&ec_encoder, &ec_encoder,
enc->initial_state, enc->initial_state,
dred_state_quant_scales_q8 + state_qoffset, dred_state_quant_scales_q8 + state_qoffset,
dred_state_dead_zone_q10 + state_qoffset, dred_state_dead_zone_q8 + state_qoffset,
dred_state_r_q8 + state_qoffset, dred_state_r_q8 + state_qoffset,
dred_state_p0_q8 + state_qoffset, dred_state_p0_q8 + state_qoffset,
DRED_STATE_DIM); DRED_STATE_DIM);
@ -291,7 +291,7 @@ int dred_encode_silk_frame(const DREDEnc *enc, unsigned char *buf, int max_chunk
&ec_encoder, &ec_encoder,
enc->latents_buffer + (i+enc->latent_offset) * DRED_LATENT_DIM, enc->latents_buffer + (i+enc->latent_offset) * DRED_LATENT_DIM,
dred_latent_quant_scales_q8 + offset, dred_latent_quant_scales_q8 + offset,
dred_latent_dead_zone_q10 + offset, dred_latent_dead_zone_q8 + offset,
dred_latent_r_q8 + offset, dred_latent_r_q8 + offset,
dred_latent_p0_q8 + offset, dred_latent_p0_q8 + offset,
DRED_LATENT_DIM DRED_LATENT_DIM