llama, mtmd: follow the reference precision of ggml_prec_set_acc calls (#31851)

* llama, mtmd: follow the reference precision of ggml_prec_set_acc calls

Requests that the reference implementation does not keep in F32 move
to GGML_PREC_BF16: the GLM4 family down and output projections, the
MiMoVL vision matmuls and the Qwen3-TTS speaker encoder conv1d. The
DeepSeek V4/V4.1 routers and the MiniMax-M3 indexer stay in F32 with
a link to the reference line that computes them in F32.

* ggml: handle GGML_PREC_BF16 the same way as GGML_PREC_F32 in backends

Backends that reject GGML_PREC_F32 on the src1 of MUL_MAT_ID now also
reject GGML_PREC_BF16 (CUDA, Hexagon, OpenCL, WebGPU, Vulkan), and
backends that accept GGML_PREC_F32 for flash attention now also accept
GGML_PREC_BF16 with the same F32 accumulation (CPU, SpaceMIT, ET).
Vulkan keeps the F32 transfer between devices for both. The CUDA
cuBLAS path already handles GGML_PREC_BF16 and is unchanged.

* tests: cover GGML_PREC_BF16 in the flash attention tests

The flash attention cases run with GGML_PREC_BF16 next to F32 and
DEFAULT, and var_to_str prints bf16 so these cases keep their own name.

* llama: request GGML_PREC_BF16 for the KQ multiplication

* cont : add TODO in clip.cpp

---------

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
This commit is contained in:
PascalandGeorgi Gerganov authored and GitHub committed 2026-11-08 08:37:46 +01:00
1 parent a2b301831d
commit d6be065572
17 files changed
+32 -27

No files matched your search

+7 -3
View File
@@ -442,7 +442,11 @@ static std::string var_to_str(ggml_type type) {
}
static std::string var_to_str(ggml_prec prec) {
return prec == GGML_PREC_F32 ? "f32" : "def";
switch (prec) {
case GGML_PREC_F32: return "f32";
case GGML_PREC_BF16: return "bf16";
default: return "def";
}
}
static std::string var_to_str(ggml_op_pool pool) {
@@ -12584,8 +12588,8 @@ static std::vector<std::unique_ptr<test_case>> make_test_cases_eval() {
for (int kv : { 113, 512, 1024, }) {
if (nr2 != 1 && kv != 512) continue;
for (int nb : { 1, 3, 32, 75, }) {
for (ggml_prec prec : {GGML_PREC_F32, GGML_PREC_DEFAULT}) {
if (hsk != 128 && prec == GGML_PREC_DEFAULT) continue;
for (ggml_prec prec : {GGML_PREC_F32, GGML_PREC_BF16, GGML_PREC_DEFAULT}) {
if (hsk != 128 && prec != GGML_PREC_F32) continue;
for (ggml_type type_KV : {GGML_TYPE_F32, GGML_TYPE_F16, GGML_TYPE_BF16, GGML_TYPE_Q8_0, GGML_TYPE_Q5_1, GGML_TYPE_Q5_0, GGML_TYPE_Q4_1, GGML_TYPE_Q4_0, GGML_TYPE_IQ4_NL}) {
if (type_KV != GGML_TYPE_F16 && hsk != 64 && hsk != 72) continue;
// DeepSeek MLA: the V cache is a sub-view of the K cache