Skip to content

mtmd: refactor audio preprocessing #17978

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
ngxson wants to merge 7 commits into
base: master
Choose a base branch
from
+387 −602
Open

mtmd: refactor audio preprocessing #17978

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
7cc2cf1
mtmd: refactor audio preprocessing
ngxson Dec 12, 2025
9e2cd84
refactor
ngxson Dec 13, 2025
cea1a90
wip
ngxson Dec 13, 2025
7b578b5
wip (2)
ngxson Dec 13, 2025
4b63e61
improve constructor
ngxson Dec 13, 2025
93290e5
fix use_natural_log
ngxson Dec 13, 2025
1aaec3b
fix padding for short input
ngxson Dec 13, 2025
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
9 changes: 9 additions & 0 deletions tools/mtmd/clip-model.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -65,6 +65,13 @@ struct clip_hparams {
int32_t n_mel_bins = 0; // whisper preprocessor
int32_t proj_stack_factor = 0; // ultravox

// audio-to-mel preprocessor params
int32_t audio_chunk_len = -1; // in seconds
int32_t audio_sample_rate = -1;
int32_t audio_n_fft = -1;
int32_t audio_window_len = -1;
int32_t audio_hop_len = -1;

// legacy
bool has_llava_projector = false;
int minicpmv_version = 0;
Expand Down Expand Up @@ -277,3 +284,5 @@ struct clip_model {
|| proj_type == PROJECTOR_TYPE_VOXTRAL;
}
};

const clip_hparams * clip_get_hparams(const struct clip_ctx * ctx);
19 changes: 16 additions & 3 deletions tools/mtmd/clip.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1141,11 +1141,15 @@ struct clip_model_loader {
bool require_stack = model.proj_type == PROJECTOR_TYPE_ULTRAVOX ||
model.proj_type == PROJECTOR_TYPE_VOXTRAL;
get_u32(KEY_A_PROJ_STACK_FACTOR, hparams.proj_stack_factor, require_stack);
if (hparams.n_mel_bins != 128) {
throw std::runtime_error(string_format("%s: only 128 mel bins are supported for ultravox\n", __func__));
}
hparams.ffn_op = FFN_GELU_ERF;
log_ffn_op = "gelu_erf"; // temporary solution for logging

// audio preprocessing params
hparams.audio_chunk_len = 30; // in seconds
hparams.audio_sample_rate = 16000;
hparams.audio_n_fft = 400;
hparams.audio_window_len = 400;
hparams.audio_hop_len = 160;
} break;
default:
break;
Expand Down Expand Up @@ -1183,6 +1187,11 @@ struct clip_model_loader {
LOG_INF("\n--- audio hparams ---\n");
LOG_INF("%s: n_mel_bins: %d\n", __func__, hparams.n_mel_bins);
LOG_INF("%s: proj_stack_factor: %d\n", __func__, hparams.proj_stack_factor);
LOG_INF("%s: audio_chunk_len: %d\n", __func__, hparams.audio_chunk_len);
LOG_INF("%s: audio_sample_rate: %d\n", __func__, hparams.audio_sample_rate);
LOG_INF("%s: audio_n_fft: %d\n", __func__, hparams.audio_n_fft);
LOG_INF("%s: audio_window_len: %d\n", __func__, hparams.audio_window_len);
LOG_INF("%s: audio_hop_len: %d\n", __func__, hparams.audio_hop_len);
}
LOG_INF("\n");
LOG_INF("%s: model size: %.2f MiB\n", __func__, model_size / 1024.0 / 1024.0);
Expand Down Expand Up @@ -3416,3 +3425,7 @@ void clip_image_f32_batch_add_mel(struct clip_image_f32_batch * batch, int n_mel
batch->entries.push_back(clip_image_f32_ptr(audio));
batch->is_audio = true;
}

const clip_hparams * clip_get_hparams(const struct clip_ctx * ctx) {
return &ctx->model.hparams;
}
Loading
Loading