d28961d81e
* llama : enable chunked fused GDN path * models : avoid Q and K repeats when using fused GDA * cont : fix comment Co-authored-by: Aman Gupta <amangupta052@gmail.com> * cont : fix the fix Co-authored-by: Aman Gupta <amangupta052@gmail.com> * cont : fix * metal : add GDN kernel (#20361) * metal : add Metal backend for GGML_OP_GATED_DELTA_NET Add a fused Metal kernel for the gated delta net recurrence op (#19504), enabling GPU-accelerated inference for DeltaNet-based models (Qwen3.5, etc.) on Apple Silicon. Supports both GDA (scalar gate) and KDA (per-row gate) modes with head_size 64 and 128. Unsupported configurations (head_size 32, non-contiguous tensors) gracefully fall back to CPU. Performance: Qwen3.5-0.8B Q4_K_M on M4 Max tg128: 170 -> 213 t/s (+25%) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * metal : validate contiguity of all input tensors in supports_op Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * metal : add algorithm equivalence comment for GDA decay path Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * cont : unslop + optimize * cont : clean-up --------- Co-authored-by: Paul Flynn <paul@arkavo.com> Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com> * CUDA: AR gated delta net improvements (#20391) * Add FastDiv to gated_delta_net_cuda * Shard columns across warps This reduces register pressure (avoids spill for S_v = 128) and gives the warp-scheduler more CTAs to schedule (thus hiding data-access latencies). * Remove unneded include in gated_delta_net.cu * Improve comments * Apply code-formating * Make sharding HIP-compatible 1. Use ggml_cuda_get_physical_warp_size() to determine warp size flexibly 2. Add test with partial warp to test sum reduction on CUDA * Remove fastdiv_s64, as we can treat neqk1 and rq3 as uint32_t * Rename variables * Enable GDN also for prefill, move TODO for chunked_GDN * Actually remove the TODO from 206890897546bd16602c3b79394fd5ea09ef199f * Get warp size at runtime warp_size is not known at compile time in hip host code. * Don't expose ggml_cuda_get_physical_warp_size on host --------- Co-authored-by: uvos <devnull@uvos.xyz> * llama : refactor llm_build_delta_net_base API --------- Co-authored-by: Aman Gupta <amangupta052@gmail.com> Co-authored-by: Paul Flynn <paul@arkavo.com> Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com> Co-authored-by: Oliver Simons <osimons@nvidia.com> Co-authored-by: uvos <devnull@uvos.xyz>
75 lines
2 KiB
C++
75 lines
2 KiB
C++
#pragma once
|
|
|
|
#include "ggml.h" // for ggml_log_level
|
|
|
|
#include <string>
|
|
#include <vector>
|
|
|
|
#ifdef __GNUC__
|
|
# if defined(__MINGW32__) && !defined(__clang__)
|
|
# define LLAMA_ATTRIBUTE_FORMAT(...) __attribute__((format(gnu_printf, __VA_ARGS__)))
|
|
# else
|
|
# define LLAMA_ATTRIBUTE_FORMAT(...) __attribute__((format(printf, __VA_ARGS__)))
|
|
# endif
|
|
#else
|
|
# define LLAMA_ATTRIBUTE_FORMAT(...)
|
|
#endif
|
|
|
|
//
|
|
// logging
|
|
//
|
|
|
|
LLAMA_ATTRIBUTE_FORMAT(2, 3)
|
|
void llama_log_internal (ggml_log_level level, const char * format, ...);
|
|
void llama_log_callback_default(ggml_log_level level, const char * text, void * user_data);
|
|
|
|
#define LLAMA_LOG(...) llama_log_internal(GGML_LOG_LEVEL_NONE , __VA_ARGS__)
|
|
#define LLAMA_LOG_INFO(...) llama_log_internal(GGML_LOG_LEVEL_INFO , __VA_ARGS__)
|
|
#define LLAMA_LOG_WARN(...) llama_log_internal(GGML_LOG_LEVEL_WARN , __VA_ARGS__)
|
|
#define LLAMA_LOG_ERROR(...) llama_log_internal(GGML_LOG_LEVEL_ERROR, __VA_ARGS__)
|
|
#define LLAMA_LOG_DEBUG(...) llama_log_internal(GGML_LOG_LEVEL_DEBUG, __VA_ARGS__)
|
|
#define LLAMA_LOG_CONT(...) llama_log_internal(GGML_LOG_LEVEL_CONT , __VA_ARGS__)
|
|
|
|
//
|
|
// helpers
|
|
//
|
|
|
|
template <typename T>
|
|
struct no_init {
|
|
T value;
|
|
no_init() = default;
|
|
};
|
|
|
|
struct time_meas {
|
|
time_meas(int64_t & t_acc, bool disable = false);
|
|
~time_meas();
|
|
|
|
const int64_t t_start_us;
|
|
|
|
int64_t & t_acc;
|
|
};
|
|
|
|
template <typename T>
|
|
struct buffer_view {
|
|
T * data;
|
|
size_t size = 0;
|
|
|
|
bool has_data() const {
|
|
return data && size > 0;
|
|
}
|
|
};
|
|
|
|
void replace_all(std::string & s, const std::string & search, const std::string & replace);
|
|
|
|
// TODO: rename to llama_format ?
|
|
LLAMA_ATTRIBUTE_FORMAT(1, 2)
|
|
std::string format(const char * fmt, ...);
|
|
|
|
std::string llama_format_tensor_shape(const std::vector<int64_t> & ne);
|
|
std::string llama_format_tensor_shape(const struct ggml_tensor * t);
|
|
|
|
std::string gguf_kv_to_str(const struct gguf_context * ctx_gguf, int i);
|
|
|
|
#define LLAMA_TENSOR_NAME_FATTN "__fattn__"
|
|
#define LLAMA_TENSOR_NAME_FGDN_AR "__fgdn_ar__"
|
|
#define LLAMA_TENSOR_NAME_FGDN_CH "__fgdn_ch__"
|