Skip to content

Lora example #54

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.

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 4 commits into
base: main
Choose a base branch
from
Open
Show file tree
Hide file tree
Changes from all commits
Commits
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
2 changes: 2 additions & 0 deletions program-data-separation/README.md
Original file line number Diff line number Diff line change
Expand Up @@ -27,5 +27,7 @@ To enable LoRA, we generate:

Multiple LoRA-adapted PTE files can share the same foundation weights and adding a model adapted to a new task incurs minimal binary size and runtime memory overhead.

Please take a look at [program-data-separation/cpp/lora_example](lora_example/) for a demo of the program-data separation APIs with LoRA. This example generates and runs a LoRA and a non-LoRA model that share foundation weights. At runtime, we see that memory usage does not double.

### Requirements
LoRA is currently supported on executorch main. [Please install ExecuTorch pip package from source](https://docs.pytorch.org/executorch/stable/using-executorch-building-from-source.html#install-executorch-pip-package-from-source), until executorch==1.0 is released.
53 changes: 40 additions & 13 deletions program-data-separation/cpp/CMakeLists.txt
Original file line number Diff line number Diff line change
Expand Up @@ -14,30 +14,57 @@ option(EXECUTORCH_BUILD_EXTENSION_TENSOR "" ON)
option(EXECUTORCH_BUILD_KERNELS_OPTIMIZED "" ON)
option(EXECUTORCH_BUILD_XNNPACK "" ON)

# Add ExecuTorch subdirectory
# Dependencies required for llm runner in lora demo.
if(EXECUTORCH_BUILD_LORA_DEMO)
option(EXECUTORCH_BUILD_EXTENSION_LLM "" ON)
option(EXECUTORCH_BUILD_EXTENSION_LLM_RUNNER "" ON)
option(EXECUTORCH_BUILD_KERNELS_LLM "" ON)
option(EXECUTORCH_BUILD_KERNELS_LLM_AOT "" ON)
endif()

# Add ExecuTorch subdirectory, after setting options.
add_subdirectory("executorch")

set(DEMO_SOURCES linear_example/main.cpp)
set(LINK_LIBS executorch
executorch::extensions
xnnpack_backend
# NOTE: xnnpack_backend has to go before
# kernels otherwise it doesn't get registered.
executorch::kernels
gflags
)

# Add sources and dependencies.
set(DEMO_SOURCES "")
if(EXECUTORCH_BUILD_LINEAR_DEMO)
list(APPEND DEMO_SOURCES "linear_example/main.cpp")
endif()
if(EXECUTORCH_BUILD_LORA_DEMO)
list(APPEND DEMO_SOURCES "lora_example/main.cpp")
endif()

# Create executable
add_executable(executorch_program_data_separation ${DEMO_SOURCES})

# Include directories
target_include_directories(executorch_program_data_separation PRIVATE ${CMAKE_CURRENT_SOURCE_DIR})

# Link libraries
target_link_libraries(
executorch_program_data_separation
PRIVATE executorch
extension_module_static
extension_flat_tensor
extension_tensor
xnnpack_backend
portable_ops_lib
portable_kernels
gflags
PRIVATE ${LINK_LIBS}
)

# Include directories for lora demo.
if(EXECUTORCH_BUILD_LORA_DEMO)
# Include directories
target_include_directories(executorch_program_data_separation PRIVATE
${CMAKE_CURRENT_SOURCE_DIR}
${CMAKE_CURRENT_SOURCE_DIR}/executorch/extension/llm/tokenizers/include
)
target_link_libraries(
executorch_program_data_separation
PUBLIC tokenizers::tokenizers
)
endif()

# Set output directory
set_target_properties(executorch_program_data_separation
PROPERTIES
Expand Down
2 changes: 1 addition & 1 deletion program-data-separation/cpp/executorch
Submodule executorch updated 328 files
Original file line number Diff line number Diff line change
Expand Up @@ -7,7 +7,7 @@ mkdir -p build
cd build

# Configure CMake
cmake -DCMAKE_BUILD_TYPE=Release ../..
cmake -DCMAKE_BUILD_TYPE=Release -DEXECUTORCH_BUILD_LINEAR_DEMO=True ../..

# Build the project
cmake --build . -j$(nproc)
Expand Down
130 changes: 130 additions & 0 deletions program-data-separation/cpp/lora_example/README.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,130 @@
# ExecuTorch LoRA Demo

This directory contains the C++ code for the LoRA demo. This demo showcases how to export and run models that share the same architecture without inflating binary file size or runtime memory.

Specifically, this demo walks through exporting and running a LoRA and non-LoRA llama model without duplication of shared foundation weights on disk or in memory.

1. Exporting LoRA and non-LoRA llama models, lowered to XNNPACK, with weights in a separate file.
2. Loading and running models with weights in a separate file.
3. Runtime weight sharing via XNNPACK.

## Size savings.

Size results will vary depending on the model, quantization and LoRA config. For this demo, we save ~5GB of disk space by storing weights in a separate, sharable file and ~5GB runtime memory by sharing weights at runtime through the XNNPACK weight cache. Detailed results are below.

### XNNPACK weight sharing.

The XNNPACK backend is a singleton. Weight sharing is implemented via the XNNPACK weight cache. At delegate init time, XNNPACK checks the weight cache for the weights it needs. If they don't exist, XNNPACK will fetch weights from the NamedDataMap (the API that exposes weights in a PTD file), pack them, store them in the weight cache and free the original. This means we won't keep around multiple copies of the same weights.

## Virtual environment setup.
Create and activate a Python virtual environment:
```bash
python3 -m venv .venv && source .venv/bin/activate && pip install --upgrade pip
```
Or alternatively, [install conda on your machine](https://conda.io/projects/conda/en/latest/user-guide/install/index.html)
```bash
conda create -yn executorch-ptd python=3.10.0 && conda activate executorch-ptd
```

Install dependencies:
LoRA isn't available in the 0.7.0 release of ExecuTorch. Instead, please install from source until ExecuTorch 1.0 is released.

[Install ExecuTorch pip package from source](https://docs.pytorch.org/executorch/stable/using-executorch-building-from-source.html#install-executorch-pip-package-from-source).

Currently, the LoRA changes aren't in nightlies. Once they are in, you can also install from the nightly build.
```
pip install executorch==0.8.0.devYYYYMMDD --extra-index-url https://download.pytorch.org/whl/nightly/cpu
```

## Export the model/s.
Change into the program-data-separation directory and create a directory to hold exported artifacts.
```bash
cd ~/executorch-examples/program-data-separation
mkdir models
```

Export models into the `models` directory. The first command will generated undelegated model/data files, and the second will generate XNNPACK-delegated model/data files.
```bash
sh export_lora.sh
```
Expect the files:
- llama_3_2_1B.pte
- llama_3_2_1B.ptd
- llama_3_2_1B_lora.pte
- foundation_weights.ptd
- tokenizer.model

llama_3_2_1B.ptd and foundation_weights.ptd contain the same contents, and you can remove llama_3_2_1B.ptd.
tokenizer.model is copied from the temp directory where we downloaded the HF artifacts. It will be used at runtime.

Note:
- PTE: contains the program execution logic.
- PTD: contains the constant tensors used by the PTE. This format is similar to safetensors, but relying on flatbuffer instead of json for serde.

Sample file sizes:
```
-rw-r--r-- 1 lfq users 4943000480 Aug 11 15:55 foundation.ptd
-rw-r--r-- 1 lfq users 1078636416 Aug 11 15:55 llama_3_2_1B_lora.pte
-rw-r--r-- 1 lfq users 1051324736 Aug 11 15:53 llama_3_2_1B.pte
```

Notice the lora - llama file size difference is about 27.3MB. This will change depending on the LoRA config. This demo is using the config from https://huggingface.co/lucylq/llama3_1B_lora/blob/main/adapter_config.json
```
{"r": 64, "lora_alpha": 128, "target_modules": ["q_proj", "v_proj", "o_proj"], "peft_type": "LORA", "base_model_name_or_path": "meta-llama/Llama-3.2-1B-Instruct"}
```

## Install runtime dependencies.
The ExecuTorch repository is configured as a git submodule at `~/executorch-examples/program-data-separation/cpp/executorch`. To initialize it:
```bash
cd ~/executorch-examples/
git submodule sync
git submodule update --init --recursive
```
Install dev requirements for ExecuTorch:

```bash
cd ~/executorch-examples/program-data-separation/cpp/executorch
pip install -r requirements-dev.txt
```

## Build the runtime.
Install some dependencies:
```bash
cd ~/executorch-examples/program-data-separation/cpp/executorch
sh examples/models/llama/install_requirements.sh
```

Build the executable:
```bash
cd ~/executorch-examples/program-data-separation/cpp/lora_example
sh build_example.sh
```

## Run the executable.
```bash
cd ~/executorch-examples/program-data-separation/cpp/lora_example

./build/bin/executorch_program_data_separation --lora_model_path=../../llama_3_2_1B_lora.pte --llama_model_path=../../llama_3_2_1B.pte --tokenizer_path=../../tokenizer.model --foundation_weights_path=../../foundation.ptd
```

You should see some logs showing the Resident Set Size (RSS) at various points of the execution. Some sample logs may look like this:

```
Generating with llama...
RSS after loading model: 7886.125000 MiB
RSS after prompt prefill: 7886.125000 MiB
RSS after finishing text generation: 7886.125000 MiB

Generating with lora...
RSS after loading model: 7933.523438 MiB
RSS after prompt prefill: 7933.523438 MiB
RSS after finishing text generation: 7933.523438 MiB
```
Notice the memory increase of ~47 MiB from running llama model to running lora model. You can see the difference without weight-sharing by removing the flag `-DEXECUTORCH_XNNPACK_ENABLE_WEIGHT_CACHE=True` from `build_example.sh`.

## Clean up.
```bash
rm -rf build
cd ~/executorch-examples/program-data-separation
rm -rf *.pte *.ptd tokenizer.model
```
15 changes: 15 additions & 0 deletions program-data-separation/cpp/lora_example/build_example.sh
Original file line number Diff line number Diff line change
@@ -0,0 +1,15 @@
#!/bin/bash
set -e

# Clean and create build directory if it doesn't exist
rm -rf build
mkdir -p build
cd build

# Configure CMake
cmake -DCMAKE_BUILD_TYPE=Release -DEXECUTORCH_BUILD_LORA_DEMO=True -DEXECUTORCH_XNNPACK_ENABLE_WEIGHT_CACHE=True ../..

# Build the project
cmake --build . -j$(nproc)

echo "Build complete! Executable located at: ./build/bin/executorch_program_data_separation"
129 changes: 129 additions & 0 deletions program-data-separation/cpp/lora_example/main.cpp
Original file line number Diff line number Diff line change
@@ -0,0 +1,129 @@
/*
* Copyright (c) Meta Platforms, Inc. and affiliates.
* All rights reserved.
*
* This source code is licensed under the BSD-style license found in the
* LICENSE file in the root directory of this source tree.
* @lint-ignore-every CLANGTIDY facebook-hte-Deprecated
*/

#include <memory>
#include <string>
#include <vector>

#include <gflags/gflags.h>

#include <executorch/extension/llm/runner/llm_runner_helper.h>
#include <executorch/extension/llm/runner/stats.h>
#include <executorch/extension/llm/runner/text_llm_runner.h>
#include <executorch/extension/llm/runner/text_prefiller.h>
#include <executorch/extension/llm/runner/text_token_generator.h>

#if defined(ET_USE_THREADPOOL)
#include <executorch/extension/threadpool/cpuinfo_utils.h>
#include <executorch/extension/threadpool/threadpool.h>
#endif

DEFINE_string(lora_model_path, "llama_3_2_1B_lora.pte",
"LoRA model serialized in flatbuffer format.");
DEFINE_string(llama_model_path, "llama_3_2_1B.pte",
"Model serialized in flatbuffer format.");
DEFINE_string(foundation_weights_path, "foundation.ptd",
"Foundation weights serialized in flatbuffer format.");

DEFINE_string(tokenizer_path, "tokenizer.model", "Tokenizer stuff.");

DEFINE_string(prompt, "The answer to the ultimate question is", "Prompt.");

DEFINE_double(temperature, 0,
"Temperature; Default is 0. 0 = greedy argmax sampling "
"(deterministic). Lower temperature = more deterministic");

DEFINE_int32(
seq_len, 128,
"Total number of tokens to generate (prompt + output). Defaults to "
"max_seq_len. If the number of input tokens + seq_len > max_seq_len, the "
"output will be truncated to max_seq_len tokens.");

using executorch::extension::Module;
using executorch::runtime::Error;
namespace llm = executorch::extension::llm;

namespace {
static constexpr int32_t kSpecialTokensSize = 256;
static inline std::unique_ptr<std::vector<std::string>>
_get_default_special_tokens() {
auto special_tokens =
std::make_unique<std::vector<std::string>>(std::vector<std::string>{
"<|begin_of_text|>", "<|end_of_text|>",
"<|reserved_special_token_0|>", "<|reserved_special_token_1|>",
"<|finetune_right_pad_id|>", "<|step_id|>", "<|start_header_id|>",
"<|end_header_id|>", "<|eom_id|>", "<|eot_id|>", "<|python_tag|>"});
// pad the rest of the special tokens with reserved tokens
ssize_t reserved_special_token_num = 2;
while (special_tokens->size() < kSpecialTokensSize) {
special_tokens->emplace_back("<|reserved_special_token_" +
std::to_string(reserved_special_token_num++) +
"|>");
}
return special_tokens;
}
} // namespace

int main(int argc, char *argv[]) {
ET_LOG(Info, "Running program-data separation lora example...");

gflags::ParseCommandLineFlags(&argc, &argv, true);

const char *lora_model_path = FLAGS_lora_model_path.c_str();
const char *llama_model_path = FLAGS_llama_model_path.c_str();
const char *foundation_weights_path = FLAGS_foundation_weights_path.c_str();

const char *tokenizer_path = FLAGS_tokenizer_path.c_str();
const char *prompt = FLAGS_prompt.c_str();
float temperature = FLAGS_temperature;
int32_t seq_len = 128;
int32_t cpu_threads = -1;

// Create tokenizers.
std::unique_ptr<tokenizers::Tokenizer> tokenizer1 =
llm::load_tokenizer(tokenizer_path, _get_default_special_tokens());
std::unique_ptr<tokenizers::Tokenizer> tokenizer2 =
llm::load_tokenizer(tokenizer_path, _get_default_special_tokens());

if (tokenizer1 == nullptr || tokenizer2 == nullptr) {
ET_LOG(Info,
"Failed to load %s as a Tiktoken, Sentencepiece or Llama2.c "
"tokenizer, make sure the artifact is one of these types",
tokenizer_path);
return 1;
}

// Create runners.
std::unique_ptr<llm::TextLLMRunner> llama_runner =
llm::create_text_llm_runner(llama_model_path, std::move(tokenizer1),
foundation_weights_path, temperature);
std::unique_ptr<llm::TextLLMRunner> lora_runner =
llm::create_text_llm_runner(lora_model_path, std::move(tokenizer2),
foundation_weights_path, temperature);

// Generate.
llm::GenerationConfig config{.seq_len = seq_len, .temperature = temperature};

ET_LOG(Info, "Generating with llama...");
auto error = llama_runner->generate(prompt, config);
if (error != Error::Ok) {
ET_LOG(Error, "Failed to generate with llama_runner, error code %zu.",
error);
return 1;
}

error = lora_runner->generate(prompt, config);
if (error != Error::Ok) {
ET_LOG(Error, "Failed to generate with lora_runner, error code %zu.",
error);
return 1;
}

return 0;
}
Loading