Skip to content

CUDA full GPU acceleration, KV cache in VRAM #1827

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

Merged
merged 18 commits into from
Jun 14, 2023
Merged
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
8 changes: 8 additions & 0 deletions examples/common.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -331,6 +331,12 @@ bool gpt_params_parse(int argc, char ** argv, gpt_params & params) {
}
#else
fprintf(stderr, "warning: llama.cpp was compiled without cuBLAS. It is not possible to set a tensor split.\n");
#endif // GGML_USE_CUBLAS
} else if (arg == "--low-vram" || arg == "-lv") {
#ifdef GGML_USE_CUBLAS
params.low_vram = true;
#else
fprintf(stderr, "warning: llama.cpp was compiled without cuBLAS. It is not possible to set lower vram usage.\n");
#endif // GGML_USE_CUBLAS
} else if (arg == "--no-mmap") {
params.use_mmap = false;
Expand Down Expand Up @@ -479,6 +485,7 @@ void gpt_print_usage(int /*argc*/, char ** argv, const gpt_params & params) {
fprintf(stderr, " -ts SPLIT --tensor-split SPLIT\n");
fprintf(stderr, " how to split tensors across multiple GPUs, comma-separated list of proportions, e.g. 3,1\n");
fprintf(stderr, " -mg i, --main-gpu i the GPU to use for scratch and small tensors\n" );
fprintf(stderr, " -lv, --low-vram don't allocate VRAM scratch buffer\n" );
#endif
fprintf(stderr, " --mtest compute maximum memory usage\n");
fprintf(stderr, " --export export the computation graph to 'llama.ggml'\n");
Expand Down Expand Up @@ -528,6 +535,7 @@ struct llama_context * llama_init_from_gpt_params(const gpt_params & params) {
lparams.n_gpu_layers = params.n_gpu_layers;
lparams.main_gpu = params.main_gpu;
memcpy(lparams.tensor_split, params.tensor_split, LLAMA_MAX_DEVICES*sizeof(float));
lparams.low_vram = params.low_vram;
lparams.seed = params.seed;
lparams.f16_kv = params.memory_f16;
lparams.use_mmap = params.use_mmap;
Expand Down
17 changes: 9 additions & 8 deletions examples/common.h
Original file line number Diff line number Diff line change
Expand Up @@ -21,15 +21,16 @@
int32_t get_num_physical_cores();

struct gpt_params {
int32_t seed = -1; // RNG seed
int32_t n_threads = get_num_physical_cores();
int32_t n_predict = -1; // new tokens to predict
int32_t n_ctx = 512; // context size
int32_t n_batch = 512; // batch size for prompt processing (must be >=32 to use BLAS)
int32_t n_keep = 0; // number of tokens to keep from initial prompt
int32_t n_gpu_layers = 0; // number of layers to store in VRAM
int32_t main_gpu = 0; // the GPU that is used for scratch and small tensors
int32_t seed = -1; // RNG seed
int32_t n_threads = get_num_physical_cores();
int32_t n_predict = -1; // new tokens to predict
int32_t n_ctx = 512; // context size
int32_t n_batch = 512; // batch size for prompt processing (must be >=32 to use BLAS)
int32_t n_keep = 0; // number of tokens to keep from initial prompt
int32_t n_gpu_layers = 0; // number of layers to store in VRAM
int32_t main_gpu = 0; // the GPU that is used for scratch and small tensors
float tensor_split[LLAMA_MAX_DEVICES] = {0}; // how split tensors should be distributed across GPUs
bool low_vram = 0; // if true, reduce VRAM usage at the cost of performance

// sampling parameters
std::unordered_map<llama_token, float> logit_bias; // logit bias for specific tokens
Expand Down
1 change: 1 addition & 0 deletions examples/main/README.md
Original file line number Diff line number Diff line change
Expand Up @@ -288,5 +288,6 @@ These options provide extra functionality and customization when running the LLa
- `-ngl N, --n-gpu-layers N`: When compiled with appropriate support (currently CLBlast or cuBLAS), this option allows offloading some layers to the GPU for computation. Generally results in increased performance.
- `-mg i, --main-gpu i`: When using multiple GPUs this option controls which GPU is used for small tensors for which the overhead of splitting the computation across all GPUs is not worthwhile. The GPU in question will use slightly more VRAM to store a scratch buffer for temporary results. By default GPU 0 is used. Requires cuBLAS.
- `-ts SPLIT, --tensor-split SPLIT`: When using multiple GPUs this option controls how large tensors should be split across all GPUs. `SPLIT` is a comma-separated list of non-negative values that assigns the proportion of data that each GPU should get in order. For example, "3,2" will assign 60% of the data to GPU 0 and 40% to GPU 1. By default the data is split in proportion to VRAM but this may not be optimal for performance. Requires cuBLAS.
- `-lv, --low-vram`: Do not allocate a VRAM scratch buffer for holding temporary results. Reduces VRAM usage at the cost of performance, particularly prompt processing speed. Requires cuBLAS.
- `--lora FNAME`: Apply a LoRA (Low-Rank Adaptation) adapter to the model (implies --no-mmap). This allows you to adapt the pretrained model to specific tasks or domains.
- `--lora-base FNAME`: Optional model to use as a base for the layers modified by the LoRA adapter. This flag is used in conjunction with the `--lora` flag, and specifies the base model for the adaptation.
1 change: 1 addition & 0 deletions examples/server/README.md
Original file line number Diff line number Diff line change
Expand Up @@ -289,6 +289,7 @@ Test();
- `-ngl N, --n-gpu-layers N`: When compiled with appropriate support (currently CLBlast or cuBLAS), this option allows offloading some layers to the GPU for computation. Generally results in increased performance.
- `-mg i, --main-gpu i`: When using multiple GPUs this option controls which GPU is used for small tensors for which the overhead of splitting the computation across all GPUs is not worthwhile. The GPU in question will use slightly more VRAM to store a scratch buffer for temporary results. By default GPU 0 is used. Requires cuBLAS.
- `-ts SPLIT, --tensor-split SPLIT`: When using multiple GPUs this option controls how large tensors should be split across all GPUs. `SPLIT` is a comma-separated list of non-negative values that assigns the proportion of data that each GPU should get in order. For example, "3,2" will assign 60% of the data to GPU 0 and 40% to GPU 1. By default the data is split in proportion to VRAM but this may not be optimal for performance. Requires cuBLAS.
- `-lv, --low-vram`: Do not allocate a VRAM scratch buffer for holding temporary results. Reduces VRAM usage at the cost of performance, particularly prompt processing speed. Requires cuBLAS.
- `--embedding`: Enable the embedding mode. **Completion function doesn't work in this mode**.
- `--host`: Set the hostname or ip address to listen. Default `127.0.0.1`;
- `--port`: Set the port to listen. Default: `8080`.
Expand Down
9 changes: 9 additions & 0 deletions examples/server/server.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -405,6 +405,7 @@ void server_print_usage(int /*argc*/, char **argv, const gpt_params &params)
fprintf(stderr, " how to split tensors across multiple GPUs, comma-separated list of proportions, e.g. 3,1\n");
fprintf(stderr, " how to split tensors across multiple GPUs, comma-separated list of proportions, e.g. 3,1\n");
fprintf(stderr, " -mg i, --main-gpu i the GPU to use for scratch and small tensors\n" );
fprintf(stderr, " -lv, --low-vram don't allocate VRAM scratch buffer\n" );
#endif
fprintf(stderr, " -m FNAME, --model FNAME\n");
fprintf(stderr, " model path (default: %s)\n", params.model.c_str());
Expand Down Expand Up @@ -537,6 +538,14 @@ bool server_params_parse(int argc, char **argv, server_params &sparams, gpt_para
}
#else
fprintf(stderr, "WARNING: llama.cpp was compiled without cuBLAS. It is not possible to set a tensor split.\n");
#endif // GGML_USE_CUBLAS
}
else if (arg == "--low-vram" || arg == "-lv")
{
#ifdef GGML_USE_CUBLAS
params.low_vram = true;
#else
fprintf(stderr, "warning: llama.cpp was compiled without cuBLAS. It is not possible to set lower vram usage.\n");
#endif // GGML_USE_CUBLAS
}
else if (arg == "--main-gpu" || arg == "-mg")
Expand Down
Loading