enne2/qwentts.cpp
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

codec: drop the fused streaming tail

Browse Source 
The fused mode appended the codec stream tail to the predictor frame
graph, so one compute produced both a frame's codes and its 80 ms of
audio with no host round trip. The experiment applied to max_batch 1
with a streaming synthesis only, it cost throughput against the
buffered flush that stays the default, and it kept a second frame
graph, its ring inputs and an init flag alive for that single case. It
is not worth keeping.

Remove the tail helpers, the fused graph of CodePredGraphSet, the
codec_fused field of qt_init_params, the --codec-fused flag of both
tools and the harness switch that exercised it. The predictor frame
unroll and the in graph sampler are untouched.
This commit is contained in:
Pascal
2026-08-05 18:20:08 +02:00
parent abab6b3bf3
commit 7b6ed4f6db
13 changed files with 20 additions and 344 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tools/tts-server.cpp
+1 -6
View File
@@ -52,8 +52,7 @@ static void print_usage(const char * prog) {
" --max-batch <n> Concurrent requests batched on the GPU (default: 1)\n"
" --no-fa Disable flash attention\n"
" --clamp-fp16 Clamp hidden states to FP16 range\n"
" --codec-chunk-dur <f> Codec decode chunk duration in seconds, wav responses (default: 24.0)\n"
" --codec-fused Decode each frame's audio inside the predictor graph (max-batch 1, streaming)\n",
" --codec-chunk-dur <f> Codec decode chunk duration in seconds, wav responses (default: 24.0)\n",
prog);
}
@@ -72,7 +71,6 @@ int main(int argc, char ** argv) {
server_config cfg;
bool use_fa = true;
bool clamp_fp16 = false;
bool codec_fused = false;
int max_batch = 1;
// Chunk sentinel : qt_init resolves a non positive value to the
// library default.
@@ -98,8 +96,6 @@ int main(int argc, char ** argv) {
clamp_fp16 = true;
} else if (!std::strcmp(arg, "--max-batch") && i + 1 < argc) {
max_batch = std::atoi(argv[++i]);
} else if (!std::strcmp(arg, "--codec-fused")) {
codec_fused = true;
} else if (!std::strcmp(arg, "--codec-chunk-dur") && i + 1 < argc) {
codec_chunk_dur = (float) std::atof(argv[++i]);
} else if (!std::strcmp(arg, "--help") || !std::strcmp(arg, "-h")) {
@@ -125,7 +121,6 @@ int main(int argc, char ** argv) {
iparams.clamp_fp16 = clamp_fp16;
iparams.max_batch = max_batch;
iparams.codec_chunk_sec = codec_chunk_dur;
iparams.codec_fused = codec_fused;
struct qt_context * q = qt_init(&iparams);
if (!q) {