Skip to content

fix: correct MiniMax H3 reference audio encoding #1886

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
jk212h20 wants to merge 1 commit into
base: master
Choose a base branch
from
+38 −17
Open

fix: correct MiniMax H3 reference audio encoding #1886

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
47 changes: 32 additions & 15 deletions src/model/vae/minimax_h3_audio_vae.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -154,8 +154,9 @@ namespace MiniMaxH3 {
const String2TensorStorage& tensor_storage_map = {},
const std::string prefix = "") override {
GGMLBlock::init_params(ctx, tensor_storage_map, prefix);
params["q_bias"] = ggml_new_tensor_1d(ctx, GGML_TYPE_F32, in_channels);
params["v_bias"] = ggml_new_tensor_1d(ctx, GGML_TYPE_F32, in_channels);
params["q_bias"] = ggml_new_tensor_1d(ctx, GGML_TYPE_F32, in_channels);
params["zero_k_bias"] = ggml_new_tensor_1d(ctx, GGML_TYPE_F32, in_channels);
params["v_bias"] = ggml_new_tensor_1d(ctx, GGML_TYPE_F32, in_channels);
}

ggml_tensor* forward(GGMLRunnerContext* ctx, ggml_tensor* x) {
Expand All @@ -166,7 +167,7 @@ namespace MiniMaxH3 {
return ggml_reshape_4d(ctx->ggml_ctx, bias, bias->ne[0], 1, 1, 1);
};
auto q = ggml_add(ctx->ggml_ctx, qkv[0], bias_shape(params["q_bias"]));
auto k = qkv[1];
auto k = ggml_add(ctx->ggml_ctx, qkv[1], bias_shape(params["zero_k_bias"]));
auto v = ggml_add(ctx->ggml_ctx, qkv[2], bias_shape(params["v_bias"]));

int64_t sequence = x->ne[1];
Expand Down Expand Up @@ -358,22 +359,38 @@ namespace MiniMaxH3 {
}

ggml_tensor* encode(GGMLRunnerContext* ctx, ggml_tensor* waveform) {
GGML_ASSERT(waveform->ne[1] == 2);
GGML_ASSERT(waveform->ne[1] * waveform->ne[2] * waveform->ne[3] == 2);
auto encoder = std::dynamic_pointer_cast<AudioEncoder>(blocks["encoder"]);
auto pre = std::dynamic_pointer_cast<AudioAttentionProjection>(blocks["pre_block"]);
auto mean_proj = std::dynamic_pointer_cast<LTXV::Conv1D>(blocks["mean_proj"]);

waveform = ggml_reshape_3d(ctx->ggml_ctx, waveform, waveform->ne[0], 1, waveform->ne[1]);
auto x = encoder->forward(ctx, waveform); // [B*S, 2048, T]
x = ggml_cont(ctx->ggml_ctx, ggml_permute(ctx->ggml_ctx, x, 1, 0, 2, 3));
x = pre->forward(ctx, x);
x = ggml_cont(ctx->ggml_ctx, ggml_permute(ctx->ggml_ctx, x, 1, 0, 2, 3));
auto z = mean_proj->forward(ctx, x);

auto mean = ggml_reshape_4d(ctx->ggml_ctx, params["latents_mean"], 1, kLatentChannels, 1, 1);
auto std = ggml_reshape_4d(ctx->ggml_ctx, params["latents_std"], 1, kLatentChannels, 1, 1);
z = ggml_div(ctx->ggml_ctx, ggml_sub(ctx->ggml_ctx, z, mean), std);
return ggml_cont(ctx->ggml_ctx, ggml_permute(ctx->ggml_ctx, z, 0, 2, 1, 3));
// GGML's batched conv1d storage interleaves the stream dimension
// with output channels. Subsequent layers then read stereo samples
// as adjacent feature channels. Run each mono stream independently,
// matching PyTorch's reshape(B*S, 1, samples), and concatenate only
// the completed normalized latents.
const int64_t streams = waveform->ne[2] * waveform->ne[3];
waveform = ggml_reshape_3d(ctx->ggml_ctx,
waveform,
waveform->ne[0],
1,
streams);
ggml_tensor* stereo_z = nullptr;
for (int64_t stream = 0; stream < streams; ++stream) {
auto mono = ggml_ext_slice(ctx->ggml_ctx, waveform, 2, stream, stream + 1);
auto x = encoder->forward(ctx, mono);
x = ggml_cont(ctx->ggml_ctx, ggml_permute(ctx->ggml_ctx, x, 1, 0, 2, 3));
x = pre->forward(ctx, x);
x = ggml_cont(ctx->ggml_ctx, ggml_permute(ctx->ggml_ctx, x, 1, 0, 2, 3));
auto z = mean_proj->forward(ctx, x);

auto mean = ggml_reshape_4d(ctx->ggml_ctx, params["latents_mean"], 1, kLatentChannels, 1, 1);
auto std = ggml_reshape_4d(ctx->ggml_ctx, params["latents_std"], 1, kLatentChannels, 1, 1);
z = ggml_div(ctx->ggml_ctx, ggml_sub(ctx->ggml_ctx, z, mean), std);
z = ggml_cont(ctx->ggml_ctx, ggml_permute(ctx->ggml_ctx, z, 0, 2, 1, 3));
stereo_z = stereo_z == nullptr ? z : ggml_concat(ctx->ggml_ctx, stereo_z, z, 1);
}
return stereo_z;
}

ggml_tensor* decode(GGMLRunnerContext* ctx, ggml_tensor* latent) {
Expand Down
8 changes: 6 additions & 2 deletions src/stable-diffusion.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4728,7 +4728,11 @@ static sd::Tensor<float> prepare_minimax_h3_reference_waveform(const sd_audio_t&
static_cast<long double>(audio.sample_count) * target_sample_rate / audio.sample_rate));
output_samples = std::max<uint64_t>(1, output_samples);
uint64_t padded_samples = (output_samples + 799) / 800 * 800;
sd::Tensor<float> waveform({static_cast<int64_t>(padded_samples), 2, 1, 1});
// Keep stereo streams planar for the mono-per-stream audio encoder:
// [samples, 1, stereo, batch]. This avoids flattening interleaved L/R
// storage into alternating samples when the encoder folds streams into
// its batch dimension.
sd::Tensor<float> waveform({static_cast<int64_t>(padded_samples), 1, 2, 1});

for (uint64_t i = 0; i < output_samples; ++i) {
long double source_pos = static_cast<long double>(i) * audio.sample_rate / target_sample_rate;
Expand All @@ -4739,7 +4743,7 @@ static sd::Tensor<float> prepare_minimax_h3_reference_waveform(const sd_audio_t&
uint32_t source_channel = audio.channels == 1 ? 0 : std::min<uint32_t>(channel, audio.channels - 1);
float a = audio.data[source0 * audio.channels + source_channel];
float b = audio.data[source1 * audio.channels + source_channel];
waveform.index(static_cast<int64_t>(i), channel, 0, 0) =
waveform.index(static_cast<int64_t>(i), 0, channel, 0) =
std::clamp(a + (b - a) * fraction, -1.f, 1.f);
}
}
Expand Down
Loading