llama : remove experimental stuff

This commit is contained in:
Georgi Gerganov 2023-09-14 22:52:01 +03:00
parent e343b8b4d8
commit e7e7b11455
No known key found for this signature in database
GPG Key ID: 449E073F9DC10735
2 changed files with 1 additions and 47 deletions

View File

@ -925,9 +925,7 @@ void ggml_metal_graph_compute(
nth1 = 1;
if (ne11 * ne12 < 4) {
[encoder setComputePipelineState:ctx->pipeline_mul_mat_f16_f32_1row];
//} else if (ne00 >= 128 && ne01 >= 8 && ne00%4 == 0) {
} else if (false) {
// TODO: with ggml_mul_mat_pad this kernel no longer seems to be needed
} else if (ne00 >= 128 && ne01 >= 8 && ne00%4 == 0) {
[encoder setComputePipelineState:ctx->pipeline_mul_mat_f16_f32_l4];
nrows = ne11;
} else {

View File

@ -438,50 +438,6 @@ static void ggml_graph_compute_helper(std::vector<uint8_t> & buf, ggml_cgraph *
ggml_graph_compute(graph, &plan);
}
//// EXPERIMENTAL:
////
//// faster matrix multiplications for tensors that do not have dimension 0 divisible by "pad"
//// the idea is to represent the original matrix multiplication:
////
//// Z = X @ Y
////
//// with the sum of two matrix multiplications:
////
//// Z = (X_0 @ Y_0) + (X_1 @ Y_1)
////
//// here X_0 and Y_0 are views of X and Y that have dimension 0 divisible by "pad"
//// and X_1 and Y_1 are the remaining views. X_1 and Y_1 end up being small matrices that can be processed with more
//// general-purpose kernels
////
//static struct ggml_tensor * ggml_mul_mat_pad(struct ggml_context * ctx, struct ggml_tensor * x, struct ggml_tensor * y, int pad = 32) {
////#if !defined(GGML_USE_METAL)
//// return ggml_mul_mat(ctx, x, y);
////#endif
//
// // use padding only if dimension 0 is at least 8 times larger than the padding
// // else we won't get much benefit from the optimization
// const int n_pad_req = 8;
//
// if (x->ne[0] % pad == 0 || x->ne[0] / pad < n_pad_req) {
// return ggml_mul_mat(ctx, x, y);
// }
//
// struct ggml_tensor * x_0 = ggml_view_3d(ctx, x, (x->ne[0]/pad)*pad, x->ne[1], x->ne[2], x->nb[1], x->nb[2], 0);
// struct ggml_tensor * x_1 = ggml_view_3d(ctx, x, x->ne[0]%pad, x->ne[1], x->ne[2], x->nb[1], x->nb[2], x_0->ne[0]*x_0->nb[0]);
//
// struct ggml_tensor * y_0 = ggml_view_3d(ctx, y, (y->ne[0]/pad)*pad, y->ne[1], y->ne[2], y->nb[1], y->nb[2], 0);
// struct ggml_tensor * y_1 = ggml_view_3d(ctx, y, y->ne[0]%pad, y->ne[1], y->ne[2], y->nb[1], y->nb[2], y_0->ne[0]*y_0->nb[0]);
//
// return ggml_add(ctx,
// ggml_mul_mat(ctx, x_0, y_0),
// ggml_mul_mat(ctx, x_1, y_1));
//}
//
//// TODO: check if other backends benefit from this and enable for all
//#if defined(GGML_USE_METAL)
////#define ggml_mul_mat ggml_mul_mat_pad
//#endif
//
// llama helpers
//