public_git/whisper.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/whisper.cpp.git synced 2026-09-30 19:41:14 +02:00
Code Issues Packages Projects Releases Wiki Activity

parakeet : add support for NVIDIA Parakeet (#3735)

Browse Source 
* parakeet : add support for NVIDIA Parakeet


Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
This commit is contained in:
Daniel Bevenius
2026-06-16 20:44:10 +02:00
committed by GitHub
co-authored by Georgi Gerganov
parent 3805e602d3
commit 9efddafb91
38 changed files with 8733 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files
examples/parakeet-quantize/parakeet-quantize.cpp
+230
View File
@@ -0,0 +1,230 @@
#include "ggml.h"
#include "ggml-backend.h"
#include "common-ggml.h"
#include <cassert>
#include <cstdio>
#include <cstring>
#include <fstream>
#include <string>
#include <vector>
struct parakeet_hparams {
int32_t n_vocab = 0;
int32_t n_audio_ctx = 0;
int32_t n_audio_state = 0;
int32_t n_audio_head = 0;
int32_t n_audio_layer = 0;
int32_t n_mels = 0;
int32_t ftype = 0;
int32_t n_fft = 0;
int32_t subsampling_factor = 0;
int32_t n_subsampling_channels = 0;
int32_t n_conv_kernel = 0;
int32_t n_pred_dim = 0;
int32_t n_pred_layers = 0;
int32_t n_tdt_durations = 0;
int32_t n_max_tokens = 0;
};
static bool parakeet_model_quantize(const std::string & fname_inp, const std::string & fname_out, ggml_ftype ftype) {
printf("%s: loading model from '%s'\n", __func__, fname_inp.c_str());
auto finp = std::ifstream(fname_inp, std::ios::binary);
if (!finp) {
fprintf(stderr, "%s: failed to open '%s' for reading\n", __func__, fname_inp.c_str());
return false;
}
auto fout = std::ofstream(fname_out, std::ios::binary);
if (!fout) {
fprintf(stderr, "%s: failed to open '%s' for writing\n", __func__, fname_out.c_str());
return false;
}
// magic
{
uint32_t magic;
finp.read((char *) &magic, sizeof(magic));
if (magic != GGML_FILE_MAGIC) {
fprintf(stderr, "%s: invalid model file (bad magic)\n", __func__);
return false;
}
fout.write((char *) &magic, sizeof(magic));
}
// hparams
parakeet_hparams hparams;
{
finp.read((char *) &hparams.n_vocab, sizeof(hparams.n_vocab));
finp.read((char *) &hparams.n_audio_ctx, sizeof(hparams.n_audio_ctx));
finp.read((char *) &hparams.n_audio_state, sizeof(hparams.n_audio_state));
finp.read((char *) &hparams.n_audio_head, sizeof(hparams.n_audio_head));
finp.read((char *) &hparams.n_audio_layer, sizeof(hparams.n_audio_layer));
finp.read((char *) &hparams.n_mels, sizeof(hparams.n_mels));
finp.read((char *) &hparams.ftype, sizeof(hparams.ftype));
finp.read((char *) &hparams.n_fft, sizeof(hparams.n_fft));
finp.read((char *) &hparams.subsampling_factor, sizeof(hparams.subsampling_factor));
finp.read((char *) &hparams.n_subsampling_channels, sizeof(hparams.n_subsampling_channels));
finp.read((char *) &hparams.n_conv_kernel, sizeof(hparams.n_conv_kernel));
finp.read((char *) &hparams.n_pred_dim, sizeof(hparams.n_pred_dim));
finp.read((char *) &hparams.n_pred_layers, sizeof(hparams.n_pred_layers));
finp.read((char *) &hparams.n_tdt_durations, sizeof(hparams.n_tdt_durations));
finp.read((char *) &hparams.n_max_tokens, sizeof(hparams.n_max_tokens));
const int32_t qntvr_src = hparams.ftype / GGML_QNT_VERSION_FACTOR;
const int32_t ftype_dst = GGML_QNT_VERSION * GGML_QNT_VERSION_FACTOR + ftype;
fprintf(stderr, "%s: n_vocab = %d\n", __func__, hparams.n_vocab);
fprintf(stderr, "%s: n_audio_state = %d\n", __func__, hparams.n_audio_state);
fprintf(stderr, "%s: n_audio_layer = %d\n", __func__, hparams.n_audio_layer);
fprintf(stderr, "%s: n_mels = %d\n", __func__, hparams.n_mels);
fprintf(stderr, "%s: ftype (src) = %d\n", __func__, hparams.ftype);
fprintf(stderr, "%s: qntvr (src) = %d\n", __func__, qntvr_src);
fprintf(stderr, "%s: ftype (dst) = %d\n", __func__, ftype_dst);
fprintf(stderr, "%s: qntvr (dst) = %d\n", __func__, GGML_QNT_VERSION);
fout.write((char *) &hparams.n_vocab, sizeof(hparams.n_vocab));
fout.write((char *) &hparams.n_audio_ctx, sizeof(hparams.n_audio_ctx));
fout.write((char *) &hparams.n_audio_state, sizeof(hparams.n_audio_state));
fout.write((char *) &hparams.n_audio_head, sizeof(hparams.n_audio_head));
fout.write((char *) &hparams.n_audio_layer, sizeof(hparams.n_audio_layer));
fout.write((char *) &hparams.n_mels, sizeof(hparams.n_mels));
fout.write((char *) &ftype_dst, sizeof(ftype_dst));
fout.write((char *) &hparams.n_fft, sizeof(hparams.n_fft));
fout.write((char *) &hparams.subsampling_factor, sizeof(hparams.subsampling_factor));
fout.write((char *) &hparams.n_subsampling_channels, sizeof(hparams.n_subsampling_channels));
fout.write((char *) &hparams.n_conv_kernel, sizeof(hparams.n_conv_kernel));
fout.write((char *) &hparams.n_pred_dim, sizeof(hparams.n_pred_dim));
fout.write((char *) &hparams.n_pred_layers, sizeof(hparams.n_pred_layers));
fout.write((char *) &hparams.n_tdt_durations, sizeof(hparams.n_tdt_durations));
fout.write((char *) &hparams.n_max_tokens, sizeof(hparams.n_max_tokens));
}
// mel filterbank
{
int32_t n_mel, n_fb;
finp.read((char *) &n_mel, sizeof(n_mel));
fout.write((char *) &n_mel, sizeof(n_mel));
finp.read((char *) &n_fb, sizeof(n_fb));
fout.write((char *) &n_fb, sizeof(n_fb));
const size_t n = (size_t) n_mel * n_fb;
std::vector<float> buf(n);
finp.read((char *) buf.data(), n * sizeof(float));
fout.write((char *) buf.data(), n * sizeof(float));
}
// window function
{
int32_t n_window;
finp.read((char *) &n_window, sizeof(n_window));
fout.write((char *) &n_window, sizeof(n_window));
std::vector<float> buf(n_window);
finp.read((char *) buf.data(), n_window * sizeof(float));
fout.write((char *) buf.data(), n_window * sizeof(float));
}
// TDT durations
{
std::vector<uint32_t> buf(hparams.n_tdt_durations);
finp.read((char *) buf.data(), hparams.n_tdt_durations * sizeof(uint32_t));
fout.write((char *) buf.data(), hparams.n_tdt_durations * sizeof(uint32_t));
}
// vocab
{
int32_t n_tokens;
finp.read((char *) &n_tokens, sizeof(n_tokens));
fout.write((char *) &n_tokens, sizeof(n_tokens));
for (int i = 0; i < n_tokens; ++i) {
int32_t len;
finp.read((char *) &len, sizeof(len));
fout.write((char *) &len, sizeof(len));
std::string token(len, '\0');
finp.read(&token[0], len);
fout.write(&token[0], len);
}
}
// tensors — quantize 2D weights skipping tensors that must stay F32:
// ggml_ssm_conv / ggml_conv2d_dw CUDA kernels require F32 weights.
// pos_bias_u / pos_bias_v are declared F32 in the loader.
const std::vector<std::string> to_quant = { ".*" };
std::vector<std::string> to_skip = {
// CUDA kernel constraints (ggml_ssm_conv / ggml_conv2d_dw require F32 weights)
"encoder\\.layers\\..+\\.conv\\.depthwise_conv\\.weight",
// Declared F32 in loader (pos_bias tensors)
"encoder\\.layers\\..+\\.self_attn\\.pos_bias_u",
"encoder\\.layers\\..+\\.self_attn\\.pos_bias_v",
};
// Prediction/joint tensors use n_pred_dim as their inner dimension. K-quant
// types (block size 256) cannot quantize 640 evenly, so keep them F32. For
// other types (Q8_0, Q4_0, block size 32) 640 is divisible and they can be
// quantized normally. The loader mirrors this logic at load time.
{
const ggml_type qtype = ggml_ftype_to_ggml_type(ftype);
const int32_t blck = ggml_blck_size(qtype);
if (blck > 1 && hparams.n_pred_dim % blck != 0) {
to_skip.push_back("decoder\\.prediction\\.embed\\.weight");
to_skip.push_back("decoder\\.prediction\\.dec_rnn\\.lstm\\.weight_ih_l.*");
to_skip.push_back("decoder\\.prediction\\.dec_rnn\\.lstm\\.weight_hh_l.*");
to_skip.push_back("joint\\.pred\\.weight");
to_skip.push_back("joint\\.joint_net\\.2\\.weight");
}
}
if (!ggml_common_quantize_0(finp, fout, ftype, to_quant, to_skip)) {
fprintf(stderr, "%s: failed to quantize tensors\n", __func__);
return false;
}
finp.close();
fout.close();
return true;
}
int main(int argc, char ** argv) {
ggml_backend_load_all();
if (argc != 4) {
fprintf(stderr, "usage: %s model-f32.bin model-quant.bin type\n", argv[0]);
ggml_print_ftypes(stderr);
return 1;
}
// initialise F16 lookup tables
{
struct ggml_init_params params = { 0, NULL, false };
struct ggml_context * ctx = ggml_init(params);
ggml_free(ctx);
}
const std::string fname_inp = argv[1];
const std::string fname_out = argv[2];
const ggml_ftype ftype = ggml_parse_ftype(argv[3]);
if (ftype == GGML_FTYPE_UNKNOWN) {
fprintf(stderr, "%s: invalid quantization type\n", argv[0]);
ggml_print_ftypes(stderr);
return 1;
}
const int64_t t_start_us = ggml_time_us();
if (!parakeet_model_quantize(fname_inp, fname_out, ftype)) {
fprintf(stderr, "%s: failed to quantize model from '%s'\n", argv[0], fname_inp.c_str());
return 1;
}
printf("\n%s: quantize time = %8.2f ms\n", argv[0], (ggml_time_us() - t_start_us) / 1000.0f);
printf("%s: output model = %s\n", argv[0], fname_out.c_str());
return 0;
}