Skip to content

Model: Warn users if diffusion-model tensors names are already prefixed #435

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

Open
wants to merge 3 commits into
base: master
Choose a base branch
from
Open

Model: Warn users if diffusion-model tensors names are already prefixed #435

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
6ba1542
Model: Warn user if wrong model type option is used
stduhpf Oct 7, 2024
ddb40b5
Fix printf
stduhpf Oct 7, 2024
ec5f18f
do not warn for empty prefix
stduhpf Oct 7, 2024
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
25 changes: 25 additions & 0 deletions model.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -915,11 +915,20 @@ bool ModelLoader::init_from_gguf_file(const std::string& file_path, const std::s

size_t total_size = 0;
size_t data_offset = gguf_get_data_offset(ctx_gguf_);

for (int i = 0; i < n_tensors; i++) {
std::string name = gguf_get_tensor_name(ctx_gguf_, i);
struct ggml_tensor* dummy = ggml_get_tensor(ctx_meta_, name.c_str());
size_t offset = data_offset + gguf_get_tensor_offset(ctx_gguf_, i);

if (!prefix.empty() && i == 0 && starts_with(name, prefix)) {
LOG_WARN("Tensors have built-in \"%s\" prefix.\n", prefix.c_str());
if (prefix == "model.diffusion_model.") {
// the user probably used `--diffusion-model` instead of `-m`
LOG_WARN("Try using `-m`or `--model` instead of `--diffusion-model`\n");
}
}

// LOG_DEBUG("%s", name.c_str());

TensorStorage tensor_storage(prefix + name, dummy->type, dummy->ne, ggml_n_dims(dummy), file_index, offset);
Expand Down Expand Up @@ -1000,6 +1009,7 @@ bool ModelLoader::init_from_safetensors_file(const std::string& file_path, const

nlohmann::json header_ = nlohmann::json::parse(header_buf.data());

int i = 0;
for (auto& item : header_.items()) {
std::string name = item.key();
nlohmann::json tensor_info = item.value();
Expand Down Expand Up @@ -1050,6 +1060,14 @@ bool ModelLoader::init_from_safetensors_file(const std::string& file_path, const
n_dims = 1;
}

if (!prefix.empty() && i++ == 0 && starts_with(name, prefix)) {
LOG_WARN("Tensors have built-in \"%s\" prefix.\n", prefix.c_str());
if (prefix == "model.diffusion_model.") {
// the user probably used `--diffusion-model` instead of `-m`
LOG_WARN("Try using `-m`or `--model` instead of `--diffusion-model`\n");
}
}

TensorStorage tensor_storage(prefix + name, type, ne, n_dims, file_index, ST_HEADER_SIZE_LEN + header_size_ + begin);
tensor_storage.reverse_ne();

Expand Down Expand Up @@ -1433,6 +1451,13 @@ bool ModelLoader::init_from_ckpt_file(const std::string& file_path, const std::s
{
std::string name = zip_entry_name(zip);
size_t pos = name.find("data.pkl");
if (!prefix.empty() && i == 0 && starts_with(name, prefix)) {
LOG_WARN("Tensors have built-in \"%s\" prefix.\n", prefix.c_str());
if (prefix == "model.diffusion_model.") {
// the user probably used `--diffusion-model` instead of `-m`
LOG_WARN("Try using `-m`or `--model` instead of `--diffusion-model`\n");
}
}
if (pos != std::string::npos) {
std::string dir = name.substr(0, pos);
printf("ZIP %d, name = %s, dir = %s \n", i, name.c_str(), dir.c_str());
Expand Down
Loading