Skip to content

quantize: skip tensor override when in fallback mode #14995

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
merged 2 commits into from
Jul 31, 2025
Merged

quantize: skip tensor override when in fallback mode #14995

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
392061c
Skip overriding when tensor is in fallback mode
EAddario Jul 30, 2025
b679810
Update comment
EAddario Jul 31, 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
6 changes: 3 additions & 3 deletions src/llama-quant.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -875,9 +875,10 @@ static void llama_model_quantize_impl(const std::string & fname_inp, const std::

// get more optimal quantization type based on the tensor shape, layer, etc.
if (!params->pure && ggml_is_quantized(default_type)) {
int fallback = qs.n_fallback;
new_type = llama_tensor_get_type(qs, new_type, tensor, ftype);
// unless the user specifies a type
if (params->tensor_types) {
// unless the user specifies a type, and the tensor geometry will not require fallback quantisation
if (params->tensor_types && qs.n_fallback - fallback == 0) {
const std::vector<tensor_quantization> & tensor_types = *static_cast<const std::vector<tensor_quantization> *>(params->tensor_types);
const std::string tensor_name(tensor->name);
for (const auto & [tname, qtype] : tensor_types) {
Expand All @@ -890,7 +891,6 @@ static void llama_model_quantize_impl(const std::string & fname_inp, const std::
}
}
}

if (params->token_embedding_type < GGML_TYPE_COUNT && strcmp(tensor->name, "token_embd.weight") == 0) {
new_type = params->token_embedding_type;
}
Expand Down
Loading