vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Optimize RWKV6 Operator Naming and Implement Multi-core CPU/ SYCL Acceleration (#10133)

Browse source 
* rwkv6: rename to wkv6

* rwkv6: support avx2 avx512 armv8 armv9

* rwkv6: update cuda file name

* rwkv6: rename params

* wkv on sycl

* sycl: add some ops

* sycl: Enhance OP support judgment

* wkv6: drop armv9 and tranfer to GGML style

ggml-ci

* sync : ggml

* update the function to use appropriate types

* fix define error

* Update ggml/src/ggml-cpu.c

* add appropriate asserts

* move element-wise functions outside

* put the declaration outside the loop

* rewrite to be more inline with the common pattern for distributing threads

* use recommended way GGML_TENSOR_LOCALS

---------

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
Co-authored-by: Diego Devesa <slarengh@gmail.com>
Co-authored-by: Plamen Minev <pacominev@gmail.com>
Co-authored-by: Yuri Khrustalev <ykhrustalev@users.noreply.github.com>
Co-authored-by: Meng, Hengyu <airdldl@163.com>
This commit is contained in:
Zhiyuan Li 2024-11-07 18:19:10 +11:00 • committed by GitHub
parent 5c333e0140
commit 3bcd40b3c5
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
22 changed files with 1977 additions and 1027 deletions
Download patch file Download diff file Expand all files Collapse all files

40
ggml/src/ggml-sycl/common.cpp
View file

@ -62,3 +62,43 @@ int64_t downsample_sycl_global_range(int64_t accumulate_block_num, int64_t block
}
return sycl_down_blk_size;
}
void ggml_sycl_op_flatten(ggml_backend_sycl_context & ctx, const ggml_tensor *src0,
const ggml_tensor *src1, ggml_tensor *dst,
const ggml_sycl_op_flatten_t op) try {
const int64_t nrows0 = ggml_nrows(src0);
const bool use_src1 = src1 != nullptr;
const int64_t nrows1 = use_src1 ? ggml_nrows(src1) : 1;
GGML_ASSERT(!use_src1 || src1->backend != GGML_BACKEND_TYPE_GPU_SPLIT);
GGML_ASSERT( dst->backend != GGML_BACKEND_TYPE_GPU_SPLIT);
ggml_tensor_extra_gpu * src0_extra = (ggml_tensor_extra_gpu *) src0->extra;
ggml_tensor_extra_gpu * src1_extra = use_src1 ? (ggml_tensor_extra_gpu *) src1->extra : nullptr;
ggml_tensor_extra_gpu * dst_extra = (ggml_tensor_extra_gpu *) dst->extra;
// dd = data device
float * src0_ddf = (float *) src0->data;
float * src1_ddf = use_src1 ? (float *) src1->data : nullptr;
float * dst_ddf = (float *) dst->data;
ggml_sycl_pool_alloc<float> src0_f(ctx.pool());
ggml_sycl_pool_alloc<float> src1_f(ctx.pool());
ggml_sycl_pool_alloc<float> dst_f(ctx.pool());
ggml_sycl_set_device(ctx.device);
queue_ptr main_stream = ctx.stream();
// GGML_SYCL_DEBUG("ctx.device=%d, main_stream=%p src0_on_device=%d, src1_on_device=%d, dst_on_device=%d\n",
// ctx.device, main_stream, src0_on_device, src1_on_device, dst_on_device);
// do the computation
op(ctx, src0, src1, dst, src0_ddf, src1_ddf, dst_ddf, main_stream);
// print_ggml_tensor("tensor", dst);
}
catch (sycl::exception const &exc) {
std::cerr << exc.what() << "Exception caught at file:" << __FILE__
<< ", line:" << __LINE__ << std::endl;
std::exit(1);
}