Skip to content

llama : enforce the same K and V cache types for DeepSeek V4; enable FA if V cache is quantized #25871

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

Merged
ggerganov merged 3 commits into from
Jul 31, 2026
Merged

llama : enforce the same K and V cache types for DeepSeek V4; enable FA if V cache is quantized #25871

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
f348aac
llama : enforce the same K and V cache types for DeepSeek V4; enable …
sszymczy Jul 18, 2026
8e0a652
llama : enforce the same K and V cache types for MLA models
sszymczy Jul 28, 2026
a8604f8
Merge remote-tracking branch 'upstream/master' into force-fa-quant-kv…
sszymczy Jul 28, 2026
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
21 changes: 16 additions & 5 deletions src/llama-context.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3541,6 +3541,22 @@ llama_context * llama_init_from_model(
}
}

if ((model->hparams.is_mla() || model->arch == LLM_ARCH_DEEPSEEK4) && params.type_k != params.type_v) {
LLAMA_LOG_ERROR("%s: model does not support different K (%s) and V (%s) cache types\n", __func__, ggml_type_name(params.type_k), ggml_type_name(params.type_v));
return nullptr;
}

if (ggml_is_quantized(params.type_v) && params.flash_attn_type != LLAMA_FLASH_ATTN_TYPE_ENABLED) {
if (params.flash_attn_type == LLAMA_FLASH_ATTN_TYPE_AUTO) {
LLAMA_LOG_INFO("%s: enabling flash_attn since it is required for quantized V cache\n", __func__);
params.flash_attn_type = LLAMA_FLASH_ATTN_TYPE_ENABLED;
}
if (params.flash_attn_type == LLAMA_FLASH_ATTN_TYPE_DISABLED) {
LLAMA_LOG_ERROR("%s: quantized V cache requires flash_attn to be enabled\n", __func__);
return nullptr;
}
}

if (params.flash_attn_type != LLAMA_FLASH_ATTN_TYPE_DISABLED && ggml_is_quantized(params.type_k)) {
const uint32_t blck_size = ggml_blck_size(params.type_k);
for (uint32_t il = 0; il < model->hparams.n_layer(); ++il) {
Expand All @@ -3563,11 +3579,6 @@ llama_context * llama_init_from_model(
}
}

if (ggml_is_quantized(params.type_v) && params.flash_attn_type == LLAMA_FLASH_ATTN_TYPE_DISABLED) {
LLAMA_LOG_ERROR("%s: V cache quantization requires flash_attn\n", __func__);
return nullptr;
}

if (params.pooling_type != LLAMA_POOLING_TYPE_UNSPECIFIED &&
params.pooling_type != model->hparams.pooling_type) {
//user-specified pooling-type is different from the model default
Expand Down
Loading