llama-cpp-pro 0.2.2

package/cpp/ggml.h

@@ -0,0 +1,2492 @@
+#pragma once
+
+//
+// GGML Tensor Library
+//
+// This documentation is still a work in progress.
+// If you wish some specific topics to be covered, feel free to drop a comment:
+//
+//   https://github.com/ggerganov/whisper.cpp/issues/40
+//
+// ## Overview
+//
+// This library implements:
+//
+//  - a set of tensor operations
+//  - automatic differentiation
+//  - basic optimization algorithms
+//
+// The aim of this library is to provide a minimalistic approach for various machine learning tasks. This includes,
+// but is not limited to, the following:
+//
+//  - linear regression
+//  - support vector machines
+//  - neural networks
+//
+// The library allows the user to define a certain function using the available tensor operations. This function
+// definition is represented internally via a computation graph. Each tensor operation in the function definition
+// corresponds to a node in the graph. Having the computation graph defined, the user can choose to compute the
+// function's value and/or its gradient with respect to the input variables. Optionally, the function can be optimized
+// using one of the available optimization algorithms.
+//
+// For example, here we define the function: f(x) = a*x^2 + b
+//
+//   {
+//       struct lm_ggml_init_params params = {
+//           .mem_size   = 16*1024*1024,
+//           .mem_buffer = NULL,
+//       };
+//
+//       // memory allocation happens here
+//       struct lm_ggml_context * ctx = lm_ggml_init(params);
+//
+//       struct lm_ggml_tensor * x = lm_ggml_new_tensor_1d(ctx, LM_GGML_TYPE_F32, 1);
+//
+//       lm_ggml_set_param(ctx, x); // x is an input variable
+//
+//       struct lm_ggml_tensor * a  = lm_ggml_new_tensor_1d(ctx, LM_GGML_TYPE_F32, 1);
+//       struct lm_ggml_tensor * b  = lm_ggml_new_tensor_1d(ctx, LM_GGML_TYPE_F32, 1);
+//       struct lm_ggml_tensor * x2 = lm_ggml_mul(ctx, x, x);
+//       struct lm_ggml_tensor * f  = lm_ggml_add(ctx, lm_ggml_mul(ctx, a, x2), b);
+//
+//       ...
+//   }
+//
+// Notice that the function definition above does not involve any actual computation. The computation is performed only
+// when the user explicitly requests it. For example, to compute the function's value at x = 2.0:
+//
+//   {
+//       ...
+//
+//       struct lm_ggml_cgraph * gf = lm_ggml_new_graph(ctx);
+//       lm_ggml_build_forward_expand(gf, f);
+//
+//       // set the input variable and parameter values
+//       lm_ggml_set_f32(x, 2.0f);
+//       lm_ggml_set_f32(a, 3.0f);
+//       lm_ggml_set_f32(b, 4.0f);
+//
+//       lm_ggml_graph_compute_with_ctx(ctx, &gf, n_threads);
+//
+//       printf("f = %f\n", lm_ggml_get_f32_1d(f, 0));
+//
+//       ...
+//   }
+//
+// The actual computation is performed in the lm_ggml_graph_compute() function.
+//
+// The lm_ggml_new_tensor_...() functions create new tensors. They are allocated in the memory buffer provided to the
+// lm_ggml_init() function. You have to be careful not to exceed the memory buffer size. Therefore, you have to know
+// in advance how much memory you need for your computation. Alternatively, you can allocate a large enough memory
+// and after defining the computation graph, call the lm_ggml_used_mem() function to find out how much memory was
+// actually needed.
+//
+// The lm_ggml_set_param() function marks a tensor as an input variable. This is used by the automatic
+// differentiation and optimization algorithms.
+//
+// The described approach allows to define the function graph once and then compute its forward or backward graphs
+// multiple times. All computations will use the same memory buffer allocated in the lm_ggml_init() function. This way
+// the user can avoid the memory allocation overhead at runtime.
+//
+// The library supports multi-dimensional tensors - up to 4 dimensions. The FP16 and FP32 data types are first class
+// citizens, but in theory the library can be extended to support FP8 and integer data types.
+//
+// Each tensor operation produces a new tensor. Initially the library was envisioned to support only the use of unary
+// and binary operations. Most of the available operations fall into one of these two categories. With time, it became
+// clear that the library needs to support more complex operations. The way to support these operations is not clear
+// yet, but a few examples are demonstrated in the following operations:
+//
+//   - lm_ggml_permute()
+//   - lm_ggml_conv_1d_1s()
+//   - lm_ggml_conv_1d_2s()
+//
+// For each tensor operator, the library implements a forward and backward computation function. The forward function
+// computes the output tensor value given the input tensor values. The backward function computes the adjoint of the
+// input tensors given the adjoint of the output tensor. For a detailed explanation of what this means, take a
+// calculus class, or watch the following video:
+//
+//   What is Automatic Differentiation?
+//   https://www.youtube.com/watch?v=wG_nF1awSSY
+//
+//
+// ## Tensor data (struct lm_ggml_tensor)
+//
+// The tensors are stored in memory via the lm_ggml_tensor struct. The structure provides information about the size of
+// the tensor, the data type, and the memory buffer where the tensor data is stored. Additionally, it contains
+// pointers to the "source" tensors - i.e. the tensors that were used to compute the current tensor. For example:
+//
+//   {
+//       struct lm_ggml_tensor * c = lm_ggml_add(ctx, a, b);
+//
+//       assert(c->src[0] == a);
+//       assert(c->src[1] == b);
+//   }
+//
+// The multi-dimensional tensors are stored in row-major order. The lm_ggml_tensor struct contains fields for the
+// number of elements in each dimension ("ne") as well as the number of bytes ("nb", a.k.a. stride). This allows
+// to store tensors that are not contiguous in memory, which is useful for operations such as transposition and
+// permutation. All tensor operations have to take the stride into account and not assume that the tensor is
+// contiguous in memory.
+//
+// The data of the tensor is accessed via the "data" pointer. For example:
+//
+//   {
+//       const int nx = 2;
+//       const int ny = 3;
+//
+//       struct lm_ggml_tensor * a = lm_ggml_new_tensor_2d(ctx, LM_GGML_TYPE_F32, nx, ny);
+//
+//       for (int y = 0; y < ny; y++) {
+//           for (int x = 0; x < nx; x++) {
+//               *(float *) ((char *) a->data + y*a->nb[1] + x*a->nb[0]) = x + y;
+//           }
+//       }
+//
+//       ...
+//   }
+//
+// Alternatively, there are helper functions, such as lm_ggml_get_f32_1d() and lm_ggml_set_f32_1d() that can be used.
+//
+// ## The matrix multiplication operator (lm_ggml_mul_mat)
+//
+// TODO
+//
+//
+// ## Multi-threading
+//
+// TODO
+//
+//
+// ## Overview of ggml.c
+//
+// TODO
+//
+//
+// ## SIMD optimizations
+//
+// TODO
+//
+//
+// ## Debugging ggml
+//
+// TODO
+//
+//
+
+#ifdef LM_GGML_SHARED
+#    if defined(_WIN32) && !defined(__MINGW32__)
+#        ifdef LM_GGML_BUILD
+#            define LM_GGML_API __declspec(dllexport) extern
+#        else
+#            define LM_GGML_API __declspec(dllimport) extern
+#        endif
+#    else
+#        define LM_GGML_API __attribute__ ((visibility ("default"))) extern
+#    endif
+#else
+#    define LM_GGML_API extern
+#endif
+
+// TODO: support for clang
+#ifdef __GNUC__
+#    define LM_GGML_DEPRECATED(func, hint) func __attribute__((deprecated(hint)))
+#elif defined(_MSC_VER)
+#    define LM_GGML_DEPRECATED(func, hint) __declspec(deprecated(hint)) func
+#else
+#    define LM_GGML_DEPRECATED(func, hint) func
+#endif
+
+#ifndef __GNUC__
+#    define LM_GGML_ATTRIBUTE_FORMAT(...)
+#elif defined(__MINGW32__) && !defined(__clang__)
+#    define LM_GGML_ATTRIBUTE_FORMAT(...) __attribute__((format(gnu_printf, __VA_ARGS__)))
+#else
+#    define LM_GGML_ATTRIBUTE_FORMAT(...) __attribute__((format(printf, __VA_ARGS__)))
+#endif
+
+#include <stdbool.h>
+#include <stddef.h>
+#include <stdint.h>
+#include <stdio.h>
+
+#define LM_GGML_FILE_MAGIC   0x67676d6c // "ggml"
+#define LM_GGML_FILE_VERSION 2
+
+#define LM_GGML_QNT_VERSION        2    // bump this on quantization format changes
+#define LM_GGML_QNT_VERSION_FACTOR 1000 // do not change this
+
+#define LM_GGML_MAX_DIMS           4
+#define LM_GGML_MAX_PARAMS         2048
+#define LM_GGML_MAX_SRC            10
+#define LM_GGML_MAX_N_THREADS      512
+#define LM_GGML_MAX_OP_PARAMS      64
+
+#ifndef LM_GGML_MAX_NAME
+#   define LM_GGML_MAX_NAME        64
+#endif
+
+#define LM_GGML_DEFAULT_N_THREADS  4
+#define LM_GGML_DEFAULT_GRAPH_SIZE 2048
+
+#if UINTPTR_MAX == 0xFFFFFFFF
+    #define LM_GGML_MEM_ALIGN 4
+#else
+    #define LM_GGML_MEM_ALIGN 16
+#endif
+
+#define LM_GGML_EXIT_SUCCESS 0
+#define LM_GGML_EXIT_ABORTED 1
+
+#define LM_GGML_ROPE_TYPE_NEOX   2
+#define LM_GGML_ROPE_TYPE_MROPE  8
+#define LM_GGML_ROPE_TYPE_VISION 24
+
+#define LM_GGML_MROPE_SECTIONS   4
+
+#define LM_GGML_UNUSED(x) (void)(x)
+#ifdef __CUDACC__
+template<typename... Args>
+__host__ __device__ constexpr inline void lm_ggml_unused_vars_impl(Args&&...) noexcept {}
+#define LM_GGML_UNUSED_VARS(...) lm_ggml_unused_vars_impl(__VA_ARGS__)
+#else
+#define LM_GGML_UNUSED_VARS(...) do { (void)sizeof((__VA_ARGS__, 0)); } while(0)
+#endif // __CUDACC__
+
+#define LM_GGML_PAD(x, n) (((x) + (n) - 1) & ~((n) - 1))
+
+#ifndef NDEBUG
+#   define LM_GGML_UNREACHABLE() do { fprintf(stderr, "statement should be unreachable\n"); abort(); } while(0)
+#elif defined(__GNUC__)
+#   define LM_GGML_UNREACHABLE() __builtin_unreachable()
+#elif defined(_MSC_VER)
+#   define LM_GGML_UNREACHABLE() __assume(0)
+#else
+#   define LM_GGML_UNREACHABLE() ((void) 0)
+#endif
+
+#ifdef __cplusplus
+#   define LM_GGML_NORETURN [[noreturn]]
+#elif defined(_MSC_VER)
+#   define LM_GGML_NORETURN __declspec(noreturn)
+#else
+#   define LM_GGML_NORETURN _Noreturn
+#endif
+
+#define LM_GGML_ABORT(...) lm_ggml_abort(__FILE__, __LINE__, __VA_ARGS__)
+#define LM_GGML_ASSERT(x) if (!(x)) LM_GGML_ABORT("LM_GGML_ASSERT(%s) failed", #x)
+
+// used to copy the number of elements and stride in bytes of tensors into local variables.
+// main purpose is to reduce code duplication and improve readability.
+//
+// example:
+//
+//    LM_GGML_TENSOR_LOCALS(int64_t, ne1, src1, ne);
+//    LM_GGML_TENSOR_LOCALS(size_t,  nb1, src1, nb);
+//
+#define LM_GGML_TENSOR_LOCALS_1(type, prefix, pointer, array) \
+    const type prefix##0 = (pointer)->array[0]; \
+    LM_GGML_UNUSED(prefix##0);
+#define LM_GGML_TENSOR_LOCALS_2(type, prefix, pointer, array) \
+    LM_GGML_TENSOR_LOCALS_1    (type, prefix, pointer, array) \
+    const type prefix##1 = (pointer)->array[1]; \
+    LM_GGML_UNUSED(prefix##1);
+#define LM_GGML_TENSOR_LOCALS_3(type, prefix, pointer, array) \
+    LM_GGML_TENSOR_LOCALS_2    (type, prefix, pointer, array) \
+    const type prefix##2 = (pointer)->array[2]; \
+    LM_GGML_UNUSED(prefix##2);
+#define LM_GGML_TENSOR_LOCALS(type, prefix, pointer, array) \
+    LM_GGML_TENSOR_LOCALS_3  (type, prefix, pointer, array) \
+    const type prefix##3 = (pointer)->array[3]; \
+    LM_GGML_UNUSED(prefix##3);
+
+#define LM_GGML_TENSOR_UNARY_OP_LOCALS \
+    LM_GGML_TENSOR_LOCALS(int64_t, ne0, src0, ne) \
+    LM_GGML_TENSOR_LOCALS(size_t,  nb0, src0, nb) \
+    LM_GGML_TENSOR_LOCALS(int64_t, ne,  dst,  ne) \
+    LM_GGML_TENSOR_LOCALS(size_t,  nb,  dst,  nb)
+
+#define LM_GGML_TENSOR_BINARY_OP_LOCALS \
+    LM_GGML_TENSOR_LOCALS(int64_t, ne0, src0, ne) \
+    LM_GGML_TENSOR_LOCALS(size_t,  nb0, src0, nb) \
+    LM_GGML_TENSOR_LOCALS(int64_t, ne1, src1, ne) \
+    LM_GGML_TENSOR_LOCALS(size_t,  nb1, src1, nb) \
+    LM_GGML_TENSOR_LOCALS(int64_t, ne,  dst,  ne) \
+    LM_GGML_TENSOR_LOCALS(size_t,  nb,  dst,  nb)
+
+#define LM_GGML_TENSOR_TERNARY_OP_LOCALS \
+    LM_GGML_TENSOR_LOCALS(int64_t, ne0, src0, ne) \
+    LM_GGML_TENSOR_LOCALS(size_t,  nb0, src0, nb) \
+    LM_GGML_TENSOR_LOCALS(int64_t, ne1, src1, ne) \
+    LM_GGML_TENSOR_LOCALS(size_t,  nb1, src1, nb) \
+    LM_GGML_TENSOR_LOCALS(int64_t, ne2, src2, ne) \
+    LM_GGML_TENSOR_LOCALS(size_t,  nb2, src2, nb) \
+    LM_GGML_TENSOR_LOCALS(int64_t, ne,  dst,  ne) \
+    LM_GGML_TENSOR_LOCALS(size_t,  nb,  dst,  nb)
+
+#define LM_GGML_TENSOR_BINARY_OP_LOCALS01 \
+    LM_GGML_TENSOR_LOCALS(int64_t, ne0, src0, ne) \
+    LM_GGML_TENSOR_LOCALS(size_t,  nb0, src0, nb) \
+    LM_GGML_TENSOR_LOCALS(int64_t, ne1, src1, ne) \
+    LM_GGML_TENSOR_LOCALS(size_t,  nb1, src1, nb)
+
+#ifdef  __cplusplus
+extern "C" {
+#endif
+
+    // Function type used in fatal error callbacks
+    typedef void (*lm_ggml_abort_callback_t)(const char * error_message);
+
+    // Set the abort callback (passing null will restore original abort functionality: printing a message to stdout)
+    // Returns the old callback for chaining
+    LM_GGML_API lm_ggml_abort_callback_t lm_ggml_set_abort_callback(lm_ggml_abort_callback_t callback);
+
+    LM_GGML_NORETURN LM_GGML_ATTRIBUTE_FORMAT(3, 4)
+    LM_GGML_API void lm_ggml_abort(const char * file, int line, const char * fmt, ...);
+
+    enum lm_ggml_status {
+        LM_GGML_STATUS_ALLOC_FAILED = -2,
+        LM_GGML_STATUS_FAILED = -1,
+        LM_GGML_STATUS_SUCCESS = 0,
+        LM_GGML_STATUS_ABORTED = 1,
+    };
+
+    // get lm_ggml_status name string
+    LM_GGML_API const char * lm_ggml_status_to_string(enum lm_ggml_status status);
+
+    // ieee 754-2008 half-precision float16
+    // todo: make this not an integral type
+    typedef uint16_t lm_ggml_fp16_t;
+    LM_GGML_API float       lm_ggml_fp16_to_fp32(lm_ggml_fp16_t);
+    LM_GGML_API lm_ggml_fp16_t lm_ggml_fp32_to_fp16(float);
+    LM_GGML_API void        lm_ggml_fp16_to_fp32_row(const lm_ggml_fp16_t *, float *, int64_t);
+    LM_GGML_API void        lm_ggml_fp32_to_fp16_row(const float *, lm_ggml_fp16_t *, int64_t);
+
+    // google brain half-precision bfloat16
+    typedef struct { uint16_t bits; } lm_ggml_bf16_t;
+    LM_GGML_API lm_ggml_bf16_t lm_ggml_fp32_to_bf16(float);
+    LM_GGML_API float       lm_ggml_bf16_to_fp32(lm_ggml_bf16_t);  // consider just doing << 16
+    LM_GGML_API void        lm_ggml_bf16_to_fp32_row(const lm_ggml_bf16_t *, float *, int64_t);
+    LM_GGML_API void        lm_ggml_fp32_to_bf16_row_ref(const float *, lm_ggml_bf16_t *, int64_t);
+    LM_GGML_API void        lm_ggml_fp32_to_bf16_row(const float *, lm_ggml_bf16_t *, int64_t);
+
+    struct lm_ggml_object;
+    struct lm_ggml_context;
+    struct lm_ggml_cgraph;
+
+    // NOTE: always add types at the end of the enum to keep backward compatibility
+    enum lm_ggml_type {
+        LM_GGML_TYPE_F32     = 0,
+        LM_GGML_TYPE_F16     = 1,
+        LM_GGML_TYPE_Q4_0    = 2,
+        LM_GGML_TYPE_Q4_1    = 3,
+        // LM_GGML_TYPE_Q4_2 = 4, support has been removed
+        // LM_GGML_TYPE_Q4_3 = 5, support has been removed
+        LM_GGML_TYPE_Q5_0    = 6,
+        LM_GGML_TYPE_Q5_1    = 7,
+        LM_GGML_TYPE_Q8_0    = 8,
+        LM_GGML_TYPE_Q8_1    = 9,
+        LM_GGML_TYPE_Q2_K    = 10,
+        LM_GGML_TYPE_Q3_K    = 11,
+        LM_GGML_TYPE_Q4_K    = 12,
+        LM_GGML_TYPE_Q5_K    = 13,
+        LM_GGML_TYPE_Q6_K    = 14,
+        LM_GGML_TYPE_Q8_K    = 15,
+        LM_GGML_TYPE_IQ2_XXS = 16,
+        LM_GGML_TYPE_IQ2_XS  = 17,
+        LM_GGML_TYPE_IQ3_XXS = 18,
+        LM_GGML_TYPE_IQ1_S   = 19,
+        LM_GGML_TYPE_IQ4_NL  = 20,
+        LM_GGML_TYPE_IQ3_S   = 21,
+        LM_GGML_TYPE_IQ2_S   = 22,
+        LM_GGML_TYPE_IQ4_XS  = 23,
+        LM_GGML_TYPE_I8      = 24,
+        LM_GGML_TYPE_I16     = 25,
+        LM_GGML_TYPE_I32     = 26,
+        LM_GGML_TYPE_I64     = 27,
+        LM_GGML_TYPE_F64     = 28,
+        LM_GGML_TYPE_IQ1_M   = 29,
+        LM_GGML_TYPE_BF16    = 30,
+        // LM_GGML_TYPE_Q4_0_4_4 = 31, support has been removed from gguf files
+        // LM_GGML_TYPE_Q4_0_4_8 = 32,
+        // LM_GGML_TYPE_Q4_0_8_8 = 33,
+        LM_GGML_TYPE_TQ1_0   = 34,
+        LM_GGML_TYPE_TQ2_0   = 35,
+        // LM_GGML_TYPE_IQ4_NL_4_4 = 36,
+        // LM_GGML_TYPE_IQ4_NL_4_8 = 37,
+        // LM_GGML_TYPE_IQ4_NL_8_8 = 38,
+        LM_GGML_TYPE_MXFP4   = 39, // MXFP4 (1 block)
+        LM_GGML_TYPE_COUNT   = 40,
+    };
+
+    // precision
+    enum lm_ggml_prec {
+        LM_GGML_PREC_DEFAULT =  0, // stored as lm_ggml_tensor.op_params, 0 by default
+        LM_GGML_PREC_F32     = 10,
+    };
+
+    // model file types
+    enum lm_ggml_ftype {
+        LM_GGML_FTYPE_UNKNOWN        = -1,
+        LM_GGML_FTYPE_ALL_F32        = 0,
+        LM_GGML_FTYPE_MOSTLY_F16     = 1,  // except 1d tensors
+        LM_GGML_FTYPE_MOSTLY_Q4_0    = 2,  // except 1d tensors
+        LM_GGML_FTYPE_MOSTLY_Q4_1    = 3,  // except 1d tensors
+        LM_GGML_FTYPE_MOSTLY_Q4_1_SOME_F16 = 4, // tok_embeddings.weight and output.weight are F16
+        LM_GGML_FTYPE_MOSTLY_Q8_0    = 7,  // except 1d tensors
+        LM_GGML_FTYPE_MOSTLY_Q5_0    = 8,  // except 1d tensors
+        LM_GGML_FTYPE_MOSTLY_Q5_1    = 9,  // except 1d tensors
+        LM_GGML_FTYPE_MOSTLY_Q2_K    = 10, // except 1d tensors
+        LM_GGML_FTYPE_MOSTLY_Q3_K    = 11, // except 1d tensors
+        LM_GGML_FTYPE_MOSTLY_Q4_K    = 12, // except 1d tensors
+        LM_GGML_FTYPE_MOSTLY_Q5_K    = 13, // except 1d tensors
+        LM_GGML_FTYPE_MOSTLY_Q6_K    = 14, // except 1d tensors
+        LM_GGML_FTYPE_MOSTLY_IQ2_XXS = 15, // except 1d tensors
+        LM_GGML_FTYPE_MOSTLY_IQ2_XS  = 16, // except 1d tensors
+        LM_GGML_FTYPE_MOSTLY_IQ3_XXS = 17, // except 1d tensors
+        LM_GGML_FTYPE_MOSTLY_IQ1_S   = 18, // except 1d tensors
+        LM_GGML_FTYPE_MOSTLY_IQ4_NL  = 19, // except 1d tensors
+        LM_GGML_FTYPE_MOSTLY_IQ3_S   = 20, // except 1d tensors
+        LM_GGML_FTYPE_MOSTLY_IQ2_S   = 21, // except 1d tensors
+        LM_GGML_FTYPE_MOSTLY_IQ4_XS  = 22, // except 1d tensors
+        LM_GGML_FTYPE_MOSTLY_IQ1_M   = 23, // except 1d tensors
+        LM_GGML_FTYPE_MOSTLY_BF16    = 24, // except 1d tensors
+        LM_GGML_FTYPE_MOSTLY_MXFP4   = 25, // except 1d tensors
+    };
+
+    // available tensor operations:
+    enum lm_ggml_op {
+        LM_GGML_OP_NONE = 0,
+
+        LM_GGML_OP_DUP,
+        LM_GGML_OP_ADD,
+        LM_GGML_OP_ADD_ID,
+        LM_GGML_OP_ADD1,
+        LM_GGML_OP_ACC,
+        LM_GGML_OP_SUB,
+        LM_GGML_OP_MUL,
+        LM_GGML_OP_DIV,
+        LM_GGML_OP_SQR,
+        LM_GGML_OP_SQRT,
+        LM_GGML_OP_LOG,
+        LM_GGML_OP_SIN,
+        LM_GGML_OP_COS,
+        LM_GGML_OP_SUM,
+        LM_GGML_OP_SUM_ROWS,
+        LM_GGML_OP_MEAN,
+        LM_GGML_OP_ARGMAX,
+        LM_GGML_OP_COUNT_EQUAL,
+        LM_GGML_OP_REPEAT,
+        LM_GGML_OP_REPEAT_BACK,
+        LM_GGML_OP_CONCAT,
+        LM_GGML_OP_SILU_BACK,
+        LM_GGML_OP_NORM, // normalize
+        LM_GGML_OP_RMS_NORM,
+        LM_GGML_OP_RMS_NORM_BACK,
+        LM_GGML_OP_GROUP_NORM,
+        LM_GGML_OP_L2_NORM,
+
+        LM_GGML_OP_MUL_MAT,
+        LM_GGML_OP_MUL_MAT_ID,
+        LM_GGML_OP_OUT_PROD,
+
+        LM_GGML_OP_SCALE,
+        LM_GGML_OP_SET,
+        LM_GGML_OP_CPY,
+        LM_GGML_OP_CONT,
+        LM_GGML_OP_RESHAPE,
+        LM_GGML_OP_VIEW,
+        LM_GGML_OP_PERMUTE,
+        LM_GGML_OP_TRANSPOSE,
+        LM_GGML_OP_GET_ROWS,
+        LM_GGML_OP_GET_ROWS_BACK,
+        LM_GGML_OP_SET_ROWS,
+        LM_GGML_OP_DIAG,
+        LM_GGML_OP_DIAG_MASK_INF,
+        LM_GGML_OP_DIAG_MASK_ZERO,
+        LM_GGML_OP_SOFT_MAX,
+        LM_GGML_OP_SOFT_MAX_BACK,
+        LM_GGML_OP_ROPE,
+        LM_GGML_OP_ROPE_BACK,
+        LM_GGML_OP_CLAMP,
+        LM_GGML_OP_CONV_TRANSPOSE_1D,
+        LM_GGML_OP_IM2COL,
+        LM_GGML_OP_IM2COL_BACK,
+        LM_GGML_OP_CONV_2D,
+        LM_GGML_OP_CONV_3D,
+        LM_GGML_OP_CONV_2D_DW,
+        LM_GGML_OP_CONV_TRANSPOSE_2D,
+        LM_GGML_OP_POOL_1D,
+        LM_GGML_OP_POOL_2D,
+        LM_GGML_OP_POOL_2D_BACK,
+        LM_GGML_OP_UPSCALE,
+        LM_GGML_OP_PAD,
+        LM_GGML_OP_PAD_REFLECT_1D,
+        LM_GGML_OP_ROLL,
+        LM_GGML_OP_ARANGE,
+        LM_GGML_OP_TIMESTEP_EMBEDDING,
+        LM_GGML_OP_ARGSORT,
+        LM_GGML_OP_LEAKY_RELU,
+
+        LM_GGML_OP_FLASH_ATTN_EXT,
+        LM_GGML_OP_FLASH_ATTN_BACK,
+        LM_GGML_OP_SSM_CONV,
+        LM_GGML_OP_SSM_SCAN,
+        LM_GGML_OP_WIN_PART,
+        LM_GGML_OP_WIN_UNPART,
+        LM_GGML_OP_GET_REL_POS,
+        LM_GGML_OP_ADD_REL_POS,
+        LM_GGML_OP_RWKV_WKV6,
+        LM_GGML_OP_GATED_LINEAR_ATTN,
+        LM_GGML_OP_RWKV_WKV7,
+
+        LM_GGML_OP_UNARY,
+
+        LM_GGML_OP_MAP_CUSTOM1,
+        LM_GGML_OP_MAP_CUSTOM2,
+        LM_GGML_OP_MAP_CUSTOM3,
+
+        LM_GGML_OP_CUSTOM,
+
+        LM_GGML_OP_CROSS_ENTROPY_LOSS,
+        LM_GGML_OP_CROSS_ENTROPY_LOSS_BACK,
+        LM_GGML_OP_OPT_STEP_ADAMW,
+        LM_GGML_OP_OPT_STEP_SGD,
+
+        LM_GGML_OP_GLU,
+
+        LM_GGML_OP_COUNT,
+    };
+
+    enum lm_ggml_unary_op {
+        LM_GGML_UNARY_OP_ABS,
+        LM_GGML_UNARY_OP_SGN,
+        LM_GGML_UNARY_OP_NEG,
+        LM_GGML_UNARY_OP_STEP,
+        LM_GGML_UNARY_OP_TANH,
+        LM_GGML_UNARY_OP_ELU,
+        LM_GGML_UNARY_OP_RELU,
+        LM_GGML_UNARY_OP_SIGMOID,
+        LM_GGML_UNARY_OP_GELU,
+        LM_GGML_UNARY_OP_GELU_QUICK,
+        LM_GGML_UNARY_OP_SILU,
+        LM_GGML_UNARY_OP_HARDSWISH,
+        LM_GGML_UNARY_OP_HARDSIGMOID,
+        LM_GGML_UNARY_OP_EXP,
+        LM_GGML_UNARY_OP_GELU_ERF,
+
+        LM_GGML_UNARY_OP_COUNT,
+    };
+
+    enum lm_ggml_glu_op {
+        LM_GGML_GLU_OP_REGLU,
+        LM_GGML_GLU_OP_GEGLU,
+        LM_GGML_GLU_OP_SWIGLU,
+        LM_GGML_GLU_OP_SWIGLU_OAI,
+        LM_GGML_GLU_OP_GEGLU_ERF,
+        LM_GGML_GLU_OP_GEGLU_QUICK,
+
+        LM_GGML_GLU_OP_COUNT,
+    };
+
+    enum lm_ggml_object_type {
+        LM_GGML_OBJECT_TYPE_TENSOR,
+        LM_GGML_OBJECT_TYPE_GRAPH,
+        LM_GGML_OBJECT_TYPE_WORK_BUFFER
+    };
+
+    enum lm_ggml_log_level {
+        LM_GGML_LOG_LEVEL_NONE  = 0,
+        LM_GGML_LOG_LEVEL_DEBUG = 1,
+        LM_GGML_LOG_LEVEL_INFO  = 2,
+        LM_GGML_LOG_LEVEL_WARN  = 3,
+        LM_GGML_LOG_LEVEL_ERROR = 4,
+        LM_GGML_LOG_LEVEL_CONT  = 5, // continue previous log
+    };
+
+    // this tensor...
+    enum lm_ggml_tensor_flag {
+        LM_GGML_TENSOR_FLAG_INPUT  =  1, // ...is an input for the GGML compute graph
+        LM_GGML_TENSOR_FLAG_OUTPUT =  2, // ...is an output for the GGML compute graph
+        LM_GGML_TENSOR_FLAG_PARAM  =  4, // ...contains trainable parameters
+        LM_GGML_TENSOR_FLAG_LOSS   =  8, // ...defines loss for numerical optimization (multiple loss tensors add up)
+    };
+
+    struct lm_ggml_init_params {
+        // memory pool
+        size_t mem_size;   // bytes
+        void * mem_buffer; // if NULL, memory will be allocated internally
+        bool   no_alloc;   // don't allocate memory for the tensor data
+    };
+
+    // n-dimensional tensor
+    struct lm_ggml_tensor {
+        enum lm_ggml_type type;
+
+        struct lm_ggml_backend_buffer * buffer;
+
+        int64_t ne[LM_GGML_MAX_DIMS]; // number of elements
+        size_t  nb[LM_GGML_MAX_DIMS]; // stride in bytes:
+                                   // nb[0] = lm_ggml_type_size(type)
+                                   // nb[1] = nb[0]   * (ne[0] / lm_ggml_blck_size(type)) + padding
+                                   // nb[i] = nb[i-1] * ne[i-1]
+
+        // compute data
+        enum lm_ggml_op op;
+
+        // op params - allocated as int32_t for alignment
+        int32_t op_params[LM_GGML_MAX_OP_PARAMS / sizeof(int32_t)];
+
+        int32_t flags;
+
+        struct lm_ggml_tensor * src[LM_GGML_MAX_SRC];
+
+        // source tensor and offset for views
+        struct lm_ggml_tensor * view_src;
+        size_t               view_offs;
+
+        void * data;
+
+        char name[LM_GGML_MAX_NAME];
+
+        void * extra; // extra things e.g. for ggml-cuda.cu
+
+        char padding[8];
+    };
+
+    static const size_t LM_GGML_TENSOR_SIZE = sizeof(struct lm_ggml_tensor);
+
+    // Abort callback
+    // If not NULL, called before ggml computation
+    // If it returns true, the computation is aborted
+    typedef bool (*lm_ggml_abort_callback)(void * data);
+
+
+    //
+    // GUID
+    //
+
+    // GUID types
+    typedef uint8_t lm_ggml_guid[16];
+    typedef lm_ggml_guid * lm_ggml_guid_t;
+
+    LM_GGML_API bool lm_ggml_guid_matches(lm_ggml_guid_t guid_a, lm_ggml_guid_t guid_b);
+
+    // misc
+
+    LM_GGML_API const char * lm_ggml_version(void);
+    LM_GGML_API const char * lm_ggml_commit(void);
+
+    LM_GGML_API void    lm_ggml_time_init(void); // call this once at the beginning of the program
+    LM_GGML_API int64_t lm_ggml_time_ms(void);
+    LM_GGML_API int64_t lm_ggml_time_us(void);
+    LM_GGML_API int64_t lm_ggml_cycles(void);
+    LM_GGML_API int64_t lm_ggml_cycles_per_ms(void);
+
+    // accepts a UTF-8 path, even on Windows
+    LM_GGML_API FILE *  lm_ggml_fopen(const char * fname, const char * mode);
+
+    LM_GGML_API void    lm_ggml_print_object (const struct lm_ggml_object * obj);
+    LM_GGML_API void    lm_ggml_print_objects(const struct lm_ggml_context * ctx);
+
+    LM_GGML_API int64_t lm_ggml_nelements (const struct lm_ggml_tensor * tensor);
+    LM_GGML_API int64_t lm_ggml_nrows     (const struct lm_ggml_tensor * tensor);
+    LM_GGML_API size_t  lm_ggml_nbytes    (const struct lm_ggml_tensor * tensor);
+    LM_GGML_API size_t  lm_ggml_nbytes_pad(const struct lm_ggml_tensor * tensor); // same as lm_ggml_nbytes() but padded to LM_GGML_MEM_ALIGN
+
+    LM_GGML_API int64_t lm_ggml_blck_size(enum lm_ggml_type type);
+    LM_GGML_API size_t  lm_ggml_type_size(enum lm_ggml_type type);             // size in bytes for all elements in a block
+    LM_GGML_API size_t  lm_ggml_row_size (enum lm_ggml_type type, int64_t ne); // size in bytes for all elements in a row
+
+    LM_GGML_DEPRECATED(
+    LM_GGML_API double lm_ggml_type_sizef(enum lm_ggml_type type), // lm_ggml_type_size()/lm_ggml_blck_size() as float
+    "use lm_ggml_row_size() instead");
+
+    LM_GGML_API const char * lm_ggml_type_name(enum lm_ggml_type type);
+    LM_GGML_API const char * lm_ggml_op_name  (enum lm_ggml_op   op);
+    LM_GGML_API const char * lm_ggml_op_symbol(enum lm_ggml_op   op);
+
+    LM_GGML_API const char * lm_ggml_unary_op_name(enum lm_ggml_unary_op op);
+    LM_GGML_API const char * lm_ggml_glu_op_name(enum lm_ggml_glu_op op);
+    LM_GGML_API const char * lm_ggml_op_desc(const struct lm_ggml_tensor * t); // unary or op name
+
+    LM_GGML_API size_t  lm_ggml_element_size(const struct lm_ggml_tensor * tensor);
+
+    LM_GGML_API bool    lm_ggml_is_quantized(enum lm_ggml_type type);
+
+    // TODO: temporary until model loading of ggml examples is refactored
+    LM_GGML_API enum lm_ggml_type lm_ggml_ftype_to_lm_ggml_type(enum lm_ggml_ftype ftype);
+
+    LM_GGML_API bool lm_ggml_is_transposed(const struct lm_ggml_tensor * tensor);
+    LM_GGML_API bool lm_ggml_is_permuted  (const struct lm_ggml_tensor * tensor);
+    LM_GGML_API bool lm_ggml_is_empty     (const struct lm_ggml_tensor * tensor);
+    LM_GGML_API bool lm_ggml_is_scalar    (const struct lm_ggml_tensor * tensor);
+    LM_GGML_API bool lm_ggml_is_vector    (const struct lm_ggml_tensor * tensor);
+    LM_GGML_API bool lm_ggml_is_matrix    (const struct lm_ggml_tensor * tensor);
+    LM_GGML_API bool lm_ggml_is_3d        (const struct lm_ggml_tensor * tensor);
+    LM_GGML_API int  lm_ggml_n_dims       (const struct lm_ggml_tensor * tensor); // returns 1 for scalars
+
+    // returns whether the tensor elements can be iterated over with a flattened index (no gaps, no permutation)
+    LM_GGML_API bool lm_ggml_is_contiguous  (const struct lm_ggml_tensor * tensor);
+    LM_GGML_API bool lm_ggml_is_contiguous_0(const struct lm_ggml_tensor * tensor); // same as lm_ggml_is_contiguous()
+    LM_GGML_API bool lm_ggml_is_contiguous_1(const struct lm_ggml_tensor * tensor); // contiguous for dims >= 1
+    LM_GGML_API bool lm_ggml_is_contiguous_2(const struct lm_ggml_tensor * tensor); // contiguous for dims >= 2
+
+    // returns whether the tensor elements are allocated as one contiguous block of memory (no gaps, but permutation ok)
+    LM_GGML_API bool lm_ggml_is_contiguously_allocated(const struct lm_ggml_tensor * tensor);
+
+    // true for tensor that is stored in memory as CxWxHxN and has been permuted to WxHxCxN
+    LM_GGML_API bool lm_ggml_is_contiguous_channels(const struct lm_ggml_tensor * tensor);
+
+    // true if the elements in dimension 0 are contiguous, or there is just 1 block of elements
+    LM_GGML_API bool lm_ggml_is_contiguous_rows(const struct lm_ggml_tensor * tensor);
+
+    LM_GGML_API bool lm_ggml_are_same_shape (const struct lm_ggml_tensor * t0, const struct lm_ggml_tensor * t1);
+    LM_GGML_API bool lm_ggml_are_same_stride(const struct lm_ggml_tensor * t0, const struct lm_ggml_tensor * t1);
+
+    LM_GGML_API bool lm_ggml_can_repeat(const struct lm_ggml_tensor * t0, const struct lm_ggml_tensor * t1);
+
+    // use this to compute the memory overhead of a tensor
+    LM_GGML_API size_t lm_ggml_tensor_overhead(void);
+
+    LM_GGML_API bool lm_ggml_validate_row_data(enum lm_ggml_type type, const void * data, size_t nbytes);
+
+    // main
+
+    LM_GGML_API struct lm_ggml_context * lm_ggml_init (struct lm_ggml_init_params params);
+    LM_GGML_API void                  lm_ggml_reset(struct lm_ggml_context * ctx);
+    LM_GGML_API void                  lm_ggml_free (struct lm_ggml_context * ctx);
+
+    LM_GGML_API size_t  lm_ggml_used_mem(const struct lm_ggml_context * ctx);
+
+    LM_GGML_API bool    lm_ggml_get_no_alloc(struct lm_ggml_context * ctx);
+    LM_GGML_API void    lm_ggml_set_no_alloc(struct lm_ggml_context * ctx, bool no_alloc);
+
+    LM_GGML_API void *  lm_ggml_get_mem_buffer     (const struct lm_ggml_context * ctx);
+    LM_GGML_API size_t  lm_ggml_get_mem_size       (const struct lm_ggml_context * ctx);
+    LM_GGML_API size_t  lm_ggml_get_max_tensor_size(const struct lm_ggml_context * ctx);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_new_tensor(
+            struct lm_ggml_context * ctx,
+            enum   lm_ggml_type type,
+            int    n_dims,
+            const int64_t *ne);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_new_tensor_1d(
+            struct lm_ggml_context * ctx,
+            enum   lm_ggml_type type,
+            int64_t ne0);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_new_tensor_2d(
+            struct lm_ggml_context * ctx,
+            enum   lm_ggml_type type,
+            int64_t ne0,
+            int64_t ne1);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_new_tensor_3d(
+            struct lm_ggml_context * ctx,
+            enum   lm_ggml_type type,
+            int64_t ne0,
+            int64_t ne1,
+            int64_t ne2);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_new_tensor_4d(
+            struct lm_ggml_context * ctx,
+            enum   lm_ggml_type type,
+            int64_t ne0,
+            int64_t ne1,
+            int64_t ne2,
+            int64_t ne3);
+
+    LM_GGML_API void * lm_ggml_new_buffer(struct lm_ggml_context * ctx, size_t nbytes);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_dup_tensor (struct lm_ggml_context * ctx, const struct lm_ggml_tensor * src);
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_view_tensor(struct lm_ggml_context * ctx, struct lm_ggml_tensor * src);
+
+    // Context tensor enumeration and lookup
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_get_first_tensor(const struct lm_ggml_context * ctx);
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_get_next_tensor (const struct lm_ggml_context * ctx, struct lm_ggml_tensor * tensor);
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_get_tensor(struct lm_ggml_context * ctx, const char * name);
+
+    // Converts a flat index into coordinates
+    LM_GGML_API void lm_ggml_unravel_index(const struct lm_ggml_tensor * tensor, int64_t i, int64_t * i0, int64_t * i1, int64_t * i2, int64_t * i3);
+
+    LM_GGML_API enum lm_ggml_unary_op lm_ggml_get_unary_op(const struct lm_ggml_tensor * tensor);
+    LM_GGML_API enum lm_ggml_glu_op lm_ggml_get_glu_op(const struct lm_ggml_tensor * tensor);
+
+    LM_GGML_API void *  lm_ggml_get_data    (const struct lm_ggml_tensor * tensor);
+    LM_GGML_API float * lm_ggml_get_data_f32(const struct lm_ggml_tensor * tensor);
+
+    LM_GGML_API const char *         lm_ggml_get_name   (const struct lm_ggml_tensor * tensor);
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_set_name   (      struct lm_ggml_tensor * tensor, const char * name);
+    LM_GGML_ATTRIBUTE_FORMAT(2, 3)
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_format_name(      struct lm_ggml_tensor * tensor, const char * fmt, ...);
+
+    // Tensor flags
+    LM_GGML_API void lm_ggml_set_input(struct lm_ggml_tensor * tensor);
+    LM_GGML_API void lm_ggml_set_output(struct lm_ggml_tensor * tensor);
+    LM_GGML_API void lm_ggml_set_param(struct lm_ggml_tensor * tensor);
+    LM_GGML_API void lm_ggml_set_loss(struct lm_ggml_tensor * tensor);
+
+    //
+    // operations on tensors with backpropagation
+    //
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_dup(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    // in-place, returns view(a)
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_dup_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_add(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_add_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_add_cast(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            enum   lm_ggml_type      type);
+
+    // dst[i0, i1, i2] = a[i0, i1, i2] + b[i0, ids[i1, i2]]
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_add_id(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            struct lm_ggml_tensor  * ids);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_add1(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_add1_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b);
+
+    // dst = a
+    // view(dst, nb1, nb2, nb3, offset) += b
+    // return dst
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_acc(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            size_t                nb1,
+            size_t                nb2,
+            size_t                nb3,
+            size_t                offset);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_acc_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            size_t                nb1,
+            size_t                nb2,
+            size_t                nb3,
+            size_t                offset);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_sub(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_sub_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_mul(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_mul_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_div(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_div_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_sqr(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_sqr_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_sqrt(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_sqrt_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_log(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_log_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_sin(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_sin_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_cos(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_cos_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    // return scalar
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_sum(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    // sums along rows, with input shape [a,b,c,d] return shape [1,b,c,d]
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_sum_rows(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    // mean along rows
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_mean(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    // argmax along rows
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_argmax(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    // count number of equal elements in a and b
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_count_equal(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b);
+
+    // if a is the same shape as b, and a is not parameter, return a
+    // otherwise, return a new tensor: repeat(a) to fit in b
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_repeat(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b);
+
+    // repeat a to the specified shape
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_repeat_4d(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+                       int64_t    ne0,
+                       int64_t    ne1,
+                       int64_t    ne2,
+                       int64_t    ne3);
+
+    // sums repetitions in a into shape of b
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_repeat_back(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b); // sum up values that are adjacent in dims > 0 instead of repeated with same stride
+
+    // concat a and b along dim
+    // used in stable-diffusion
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_concat(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            int                   dim);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_abs(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_abs_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_sgn(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_sgn_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_neg(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_neg_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_step(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_step_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_tanh(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_tanh_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_elu(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_elu_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_relu(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_leaky_relu(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a, float negative_slope, bool inplace);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_relu_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_sigmoid(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_sigmoid_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_gelu(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_gelu_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    // GELU using erf (error function) when possible
+    // some backends may fallback to approximation based on Abramowitz and Stegun formula
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_gelu_erf(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_gelu_erf_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_gelu_quick(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_gelu_quick_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_silu(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_silu_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    // a - x
+    // b - dy
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_silu_back(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b);
+
+    // hardswish(x) = x * relu6(x + 3) / 6
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_hardswish(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    // hardsigmoid(x) = relu6(x + 3) / 6
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_hardsigmoid(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_exp(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_exp_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    // gated linear unit ops
+    // A: n columns, r rows,
+    // result is n / 2 columns, r rows,
+    // expects gate in second half of row, unless swapped is true
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_glu(
+            struct lm_ggml_context * ctx,
+             struct lm_ggml_tensor * a,
+             enum lm_ggml_glu_op     op,
+             bool                 swapped);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_reglu(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_reglu_swapped(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_geglu(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_geglu_swapped(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_swiglu(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_swiglu_swapped(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_geglu_erf(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_geglu_erf_swapped(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_geglu_quick(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_geglu_quick_swapped(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    // A: n columns, r rows,
+    // B: n columns, r rows,
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_glu_split(
+            struct lm_ggml_context * ctx,
+             struct lm_ggml_tensor * a,
+             struct lm_ggml_tensor * b,
+             enum lm_ggml_glu_op     op);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_reglu_split(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_geglu_split(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_swiglu_split(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_geglu_erf_split(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_geglu_quick_split(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_swiglu_oai(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            float                 alpha,
+            float                 limit);
+
+    // normalize along rows
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_norm(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            float                 eps);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_norm_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            float                 eps);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_rms_norm(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            float                 eps);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_rms_norm_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            float                 eps);
+
+    // group normalize along ne0*ne1*n_groups
+    // used in stable-diffusion
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_group_norm(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int                   n_groups,
+            float                 eps);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_group_norm_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int                   n_groups,
+            float                 eps);
+
+    // l2 normalize along rows
+    // used in rwkv v7
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_l2_norm(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            float                 eps);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_l2_norm_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            float                 eps);
+
+    // a - x
+    // b - dy
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_rms_norm_back(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            float                 eps);
+
+    // A: k columns, n rows => [ne03, ne02, n, k]
+    // B: k columns, m rows  (i.e. we transpose it internally) => [ne03 * x, ne02 * y, m, k]
+    // result is n columns, m rows => [ne03 * x, ne02 * y, m, n]
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_mul_mat(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b);
+
+    // change the precision of a matrix multiplication
+    // set to LM_GGML_PREC_F32 for higher precision (useful for phi-2)
+    LM_GGML_API void lm_ggml_mul_mat_set_prec(
+            struct lm_ggml_tensor * a,
+            enum lm_ggml_prec       prec);
+
+    // indirect matrix multiplication
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_mul_mat_id(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * as,
+            struct lm_ggml_tensor  * b,
+            struct lm_ggml_tensor  * ids);
+
+    // A: m columns, n rows,
+    // B: p columns, n rows,
+    // result is m columns, p rows
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_out_prod(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b);
+
+    //
+    // operations on tensors without backpropagation
+    //
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_scale(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            float                 s);
+
+    // in-place, returns view(a)
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_scale_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            float                 s);
+
+    // x = s * a + b
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_scale_bias(
+        struct lm_ggml_context * ctx,
+        struct lm_ggml_tensor  * a,
+        float                 s,
+        float                 b);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_scale_bias_inplace(
+        struct lm_ggml_context * ctx,
+        struct lm_ggml_tensor  * a,
+        float                 s,
+        float                 b);
+
+    // b -> view(a,offset,nb1,nb2,3), return modified a
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_set(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            size_t                nb1,
+            size_t                nb2,
+            size_t                nb3,
+            size_t                offset); // in bytes
+
+    // b -> view(a,offset,nb1,nb2,3), return view(a)
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_set_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            size_t                nb1,
+            size_t                nb2,
+            size_t                nb3,
+            size_t                offset); // in bytes
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_set_1d(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            size_t                offset); // in bytes
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_set_1d_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            size_t                offset); // in bytes
+
+    // b -> view(a,offset,nb1,nb2,3), return modified a
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_set_2d(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            size_t                nb1,
+            size_t                offset); // in bytes
+
+    // b -> view(a,offset,nb1,nb2,3), return view(a)
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_set_2d_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            size_t                nb1,
+            size_t                offset); // in bytes
+
+    // a -> b, return view(b)
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_cpy(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_cast(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            enum   lm_ggml_type      type);
+
+    // make contiguous
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_cont(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    // make contiguous, with new shape
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_cont_1d(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int64_t               ne0);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_cont_2d(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int64_t               ne0,
+            int64_t               ne1);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_cont_3d(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int64_t               ne0,
+            int64_t               ne1,
+            int64_t               ne2);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_cont_4d(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int64_t               ne0,
+            int64_t               ne1,
+            int64_t               ne2,
+            int64_t               ne3);
+
+    // return view(a), b specifies the new shape
+    // TODO: when we start computing gradient, make a copy instead of view
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_reshape(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b);
+
+    // return view(a)
+    // TODO: when we start computing gradient, make a copy instead of view
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_reshape_1d(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int64_t               ne0);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_reshape_2d(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int64_t               ne0,
+            int64_t               ne1);
+
+    // return view(a)
+    // TODO: when we start computing gradient, make a copy instead of view
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_reshape_3d(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int64_t               ne0,
+            int64_t               ne1,
+            int64_t               ne2);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_reshape_4d(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int64_t               ne0,
+            int64_t               ne1,
+            int64_t               ne2,
+            int64_t               ne3);
+
+    // offset in bytes
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_view_1d(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int64_t               ne0,
+            size_t                offset);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_view_2d(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int64_t               ne0,
+            int64_t               ne1,
+            size_t                nb1, // row stride in bytes
+            size_t                offset);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_view_3d(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int64_t               ne0,
+            int64_t               ne1,
+            int64_t               ne2,
+            size_t                nb1, // row   stride in bytes
+            size_t                nb2, // slice stride in bytes
+            size_t                offset);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_view_4d(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int64_t               ne0,
+            int64_t               ne1,
+            int64_t               ne2,
+            int64_t               ne3,
+            size_t                nb1, // row   stride in bytes
+            size_t                nb2, // slice stride in bytes
+            size_t                nb3,
+            size_t                offset);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_permute(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int                   axis0,
+            int                   axis1,
+            int                   axis2,
+            int                   axis3);
+
+    // alias for lm_ggml_permute(ctx, a, 1, 0, 2, 3)
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_transpose(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    // supports 3D: a->ne[2] == b->ne[1]
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_get_rows(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,  // data
+            struct lm_ggml_tensor  * b); // row indices
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_get_rows_back(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,  // gradients of lm_ggml_get_rows result
+            struct lm_ggml_tensor  * b,  // row indices
+            struct lm_ggml_tensor  * c); // data for lm_ggml_get_rows, only used for its shape
+
+    // a TD  [n_embd, ne1,    ne2,    ne3]
+    // b TS  [n_embd, n_rows, ne02,   ne03] | ne02 == ne2, ne03 == ne3
+    // c I64 [n_rows, ne11,   ne12,   1]    | c[i] in [0, ne1)
+    //
+    // undefined behavior if destination rows overlap
+    //
+    // broadcast:
+    //   ne2 % ne11 == 0
+    //   ne3 % ne12 == 0
+    //
+    // return view(a)
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_set_rows(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,  // destination
+            struct lm_ggml_tensor  * b,  // source
+            struct lm_ggml_tensor  * c); // row indices
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_diag(
+        struct lm_ggml_context     * ctx,
+        struct lm_ggml_tensor      * a);
+
+    // set elements above the diagonal to -INF
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_diag_mask_inf(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int                   n_past);
+
+    // in-place, returns view(a)
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_diag_mask_inf_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int                   n_past);
+
+    // set elements above the diagonal to 0
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_diag_mask_zero(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int                   n_past);
+
+    // in-place, returns view(a)
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_diag_mask_zero_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int                   n_past);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_soft_max(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    // in-place, returns view(a)
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_soft_max_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a);
+
+    // a    [ne0, ne01, ne02, ne03]
+    // mask [ne0, ne11, ne12, ne13] | ne11 >= ne01, F16 or F32, optional
+    //
+    // broadcast:
+    //   ne02 % ne12 == 0
+    //   ne03 % ne13 == 0
+    //
+    // fused soft_max(a*scale + mask*(ALiBi slope))
+    // max_bias = 0.0f for no ALiBi
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_soft_max_ext(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * mask,
+            float                 scale,
+            float                 max_bias);
+
+    LM_GGML_API void lm_ggml_soft_max_add_sinks(
+            struct lm_ggml_tensor * a,
+            struct lm_ggml_tensor * sinks);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_soft_max_ext_back(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            float                 scale,
+            float                 max_bias);
+
+    // in-place, returns view(a)
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_soft_max_ext_back_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            float                 scale,
+            float                 max_bias);
+
+    // rotary position embedding
+    // if (mode & 1) - skip n_past elements (NOT SUPPORTED)
+    // if (mode & LM_GGML_ROPE_TYPE_NEOX) - GPT-NeoX style
+    //
+    // b is an int32 vector with size a->ne[2], it contains the positions
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_rope(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            int                   n_dims,
+            int                   mode);
+
+    // in-place, returns view(a)
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_rope_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            int                   n_dims,
+            int                   mode);
+
+    // custom RoPE
+    // c is freq factors (e.g. phi3-128k), (optional)
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_rope_ext(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            struct lm_ggml_tensor  * c,
+            int                   n_dims,
+            int                   mode,
+            int                   n_ctx_orig,
+            float                 freq_base,
+            float                 freq_scale,
+            float                 ext_factor,
+            float                 attn_factor,
+            float                 beta_fast,
+            float                 beta_slow);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_rope_multi(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            struct lm_ggml_tensor  * c,
+            int                   n_dims,
+            int                   sections[LM_GGML_MROPE_SECTIONS],
+            int                   mode,
+            int                   n_ctx_orig,
+            float                 freq_base,
+            float                 freq_scale,
+            float                 ext_factor,
+            float                 attn_factor,
+            float                 beta_fast,
+            float                 beta_slow);
+
+    // in-place, returns view(a)
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_rope_ext_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            struct lm_ggml_tensor  * c,
+            int                   n_dims,
+            int                   mode,
+            int                   n_ctx_orig,
+            float                 freq_base,
+            float                 freq_scale,
+            float                 ext_factor,
+            float                 attn_factor,
+            float                 beta_fast,
+            float                 beta_slow);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_rope_multi_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            struct lm_ggml_tensor  * c,
+            int                   n_dims,
+            int                   sections[LM_GGML_MROPE_SECTIONS],
+            int                   mode,
+            int                   n_ctx_orig,
+            float                 freq_base,
+            float                 freq_scale,
+            float                 ext_factor,
+            float                 attn_factor,
+            float                 beta_fast,
+            float                 beta_slow);
+
+    LM_GGML_DEPRECATED(LM_GGML_API struct lm_ggml_tensor * lm_ggml_rope_custom(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            int                   n_dims,
+            int                   mode,
+            int                   n_ctx_orig,
+            float                 freq_base,
+            float                 freq_scale,
+            float                 ext_factor,
+            float                 attn_factor,
+            float                 beta_fast,
+            float                 beta_slow),
+        "use lm_ggml_rope_ext instead");
+
+    LM_GGML_DEPRECATED(LM_GGML_API struct lm_ggml_tensor * lm_ggml_rope_custom_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            int                   n_dims,
+            int                   mode,
+            int                   n_ctx_orig,
+            float                 freq_base,
+            float                 freq_scale,
+            float                 ext_factor,
+            float                 attn_factor,
+            float                 beta_fast,
+            float                 beta_slow),
+        "use lm_ggml_rope_ext_inplace instead");
+
+    // compute correction dims for YaRN RoPE scaling
+    LM_GGML_API void lm_ggml_rope_yarn_corr_dims(
+        int n_dims, int n_ctx_orig, float freq_base, float beta_fast, float beta_slow, float dims[2]);
+
+    // rotary position embedding backward, i.e compute dx from dy
+    // a - dy
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_rope_ext_back(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a, // gradients of lm_ggml_rope result
+            struct lm_ggml_tensor  * b, // positions
+            struct lm_ggml_tensor  * c, // freq factors
+            int                   n_dims,
+            int                   mode,
+            int                   n_ctx_orig,
+            float                 freq_base,
+            float                 freq_scale,
+            float                 ext_factor,
+            float                 attn_factor,
+            float                 beta_fast,
+            float                 beta_slow);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_rope_multi_back(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            struct lm_ggml_tensor  * c,
+            int                   n_dims,
+            int                   sections[4],
+            int                   mode,
+            int                   n_ctx_orig,
+            float                 freq_base,
+            float                 freq_scale,
+            float                 ext_factor,
+            float                 attn_factor,
+            float                 beta_fast,
+            float                 beta_slow);
+
+
+    // clamp
+    // in-place, returns view(a)
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_clamp(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            float                 min,
+            float                 max);
+
+    // im2col
+    // converts data into a format that effectively results in a convolution when combined with matrix multiplication
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_im2col(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,  // convolution kernel
+            struct lm_ggml_tensor  * b,  // data
+            int                   s0, // stride dimension 0
+            int                   s1, // stride dimension 1
+            int                   p0, // padding dimension 0
+            int                   p1, // padding dimension 1
+            int                   d0, // dilation dimension 0
+            int                   d1, // dilation dimension 1
+            bool                  is_2D,
+            enum lm_ggml_type        dst_type);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_im2col_back(
+        struct lm_ggml_context * ctx,
+        struct lm_ggml_tensor  * a,  // convolution kernel
+        struct lm_ggml_tensor  * b,  // gradient of im2col output
+        int64_t             * ne, // shape of im2col input
+        int                   s0, // stride dimension 0
+        int                   s1, // stride dimension 1
+        int                   p0, // padding dimension 0
+        int                   p1, // padding dimension 1
+        int                   d0, // dilation dimension 0
+        int                   d1, // dilation dimension 1
+        bool                  is_2D);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_conv_1d(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,   // convolution kernel
+            struct lm_ggml_tensor  * b,   // data
+            int                   s0,  // stride
+            int                   p0,  // padding
+            int                   d0); // dilation
+
+    // conv_1d with padding = half
+    // alias for lm_ggml_conv_1d(a, b, s, a->ne[0]/2, d)
+    LM_GGML_API struct lm_ggml_tensor* lm_ggml_conv_1d_ph(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,  // convolution kernel
+            struct lm_ggml_tensor  * b,  // data
+            int                   s,  // stride
+            int                   d); // dilation
+
+    // depthwise
+    // TODO: this is very likely wrong for some cases! - needs more testing
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_conv_1d_dw(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,   // convolution kernel
+            struct lm_ggml_tensor  * b,   // data
+            int                   s0,  // stride
+            int                   p0,  // padding
+            int                   d0); // dilation
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_conv_1d_dw_ph(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,   // convolution kernel
+            struct lm_ggml_tensor  * b,   // data
+            int                   s0,  // stride
+            int                   d0); // dilation
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_conv_transpose_1d(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,   // convolution kernel
+            struct lm_ggml_tensor  * b,   // data
+            int                   s0,  // stride
+            int                   p0,  // padding
+            int                   d0); // dilation
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_conv_2d(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,   // convolution kernel
+            struct lm_ggml_tensor  * b,   // data
+            int                   s0,  // stride dimension 0
+            int                   s1,  // stride dimension 1
+            int                   p0,  // padding dimension 0
+            int                   p1,  // padding dimension 1
+            int                   d0,  // dilation dimension 0
+            int                   d1); // dilation dimension 1
+
+    // kernel size is a->ne[0] x a->ne[1]
+    // stride is equal to kernel size
+    // padding is zero
+    // example:
+    // a:     16   16    3  768
+    // b:   1024 1024    3    1
+    // res:   64   64  768    1
+    // used in sam
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_conv_2d_sk_p0(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b);
+
+    // kernel size is a->ne[0] x a->ne[1]
+    // stride is 1
+    // padding is half
+    // example:
+    // a:      3    3    256  256
+    // b:     64   64    256    1
+    // res:   64   64    256    1
+    // used in sam
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_conv_2d_s1_ph(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b);
+
+    // depthwise (via im2col and mul_mat)
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_conv_2d_dw(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,  // convolution kernel
+            struct lm_ggml_tensor  * b,  // data
+            int                  s0,  // stride dimension 0
+            int                  s1,  // stride dimension 1
+            int                  p0,  // padding dimension 0
+            int                  p1,  // padding dimension 1
+            int                  d0,  // dilation dimension 0
+            int                  d1); // dilation dimension 1
+
+    // Depthwise 2D convolution
+    // may be faster than lm_ggml_conv_2d_dw, but not available in all backends
+    // a:   KW    KH    1    C    convolution kernel
+    // b:   W     H     C    N    input data
+    // res: W_out H_out C    N
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_conv_2d_dw_direct(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            int                   stride0,
+            int                   stride1,
+            int                   pad0,
+            int                   pad1,
+            int                   dilation0,
+            int                   dilation1);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_conv_transpose_2d_p0(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            int                   stride);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_conv_2d_direct(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,   // convolution kernel [KW, KH, IC, OC]
+            struct lm_ggml_tensor  * b,   // input data [W, H, C, N]
+            int                   s0,  // stride dimension 0
+            int                   s1,  // stride dimension 1
+            int                   p0,  // padding dimension 0
+            int                   p1,  // padding dimension 1
+            int                   d0,  // dilation dimension 0
+            int                   d1); // dilation dimension 1
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_conv_3d(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,   // kernel [KW, KH, KD, IC * OC]
+            struct lm_ggml_tensor  * b,   // input  [W, H, D, C * N]
+            int                   s0,  // stride
+            int                   s1,
+            int                   s2,
+            int                   p0,  // padding
+            int                   p1,
+            int                   p2,
+            int                   d0,  // dilation
+            int                   d1,
+            int                   d2,
+            int                   n_channels,
+            int                   n_batch,
+            int                   n_channels_out);
+
+    enum lm_ggml_op_pool {
+        LM_GGML_OP_POOL_MAX,
+        LM_GGML_OP_POOL_AVG,
+        LM_GGML_OP_POOL_COUNT,
+    };
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_pool_1d(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            enum lm_ggml_op_pool     op,
+            int                   k0, // kernel size
+            int                   s0, // stride
+            int                   p0); // padding
+
+    // the result will have 2*p0 padding for the first dimension
+    // and 2*p1 padding for the second dimension
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_pool_2d(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            enum lm_ggml_op_pool     op,
+            int                   k0,
+            int                   k1,
+            int                   s0,
+            int                   s1,
+            float                 p0,
+            float                 p1);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_pool_2d_back(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * af, // "a"/input used in forward pass
+            enum lm_ggml_op_pool     op,
+            int                   k0,
+            int                   k1,
+            int                   s0,
+            int                   s1,
+            float                 p0,
+            float                 p1);
+
+    enum lm_ggml_scale_mode {
+        LM_GGML_SCALE_MODE_NEAREST  = 0,
+        LM_GGML_SCALE_MODE_BILINEAR = 1,
+
+        LM_GGML_SCALE_MODE_COUNT
+    };
+
+    enum lm_ggml_scale_flag {
+        LM_GGML_SCALE_FLAG_ALIGN_CORNERS = (1 << 8)
+    };
+
+    // interpolate
+    // multiplies ne0 and ne1 by scale factor
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_upscale(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int                   scale_factor,
+            enum lm_ggml_scale_mode  mode);
+
+    // interpolate
+    // interpolate scale to specified dimensions
+    LM_GGML_DEPRECATED(LM_GGML_API struct lm_ggml_tensor * lm_ggml_upscale_ext(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int                   ne0,
+            int                   ne1,
+            int                   ne2,
+            int                   ne3,
+            enum lm_ggml_scale_mode  mode),
+        "use lm_ggml_interpolate instead");
+
+    // Up- or downsamples the input to the specified size.
+    // 2D scale modes (eg. bilinear) are applied to the first two dimensions.
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_interpolate(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int64_t               ne0,
+            int64_t               ne1,
+            int64_t               ne2,
+            int64_t               ne3,
+            uint32_t              mode); // lm_ggml_scale_mode [ | lm_ggml_scale_flag...]
+
+    // pad each dimension with zeros: [x, ..., x] -> [x, ..., x, 0, ..., 0]
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_pad(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int                  p0,
+            int                  p1,
+            int                  p2,
+            int                  p3);
+
+    // pad each dimension with reflection: [a, b, c, d] -> [b, a, b, c, d, c]
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_pad_reflect_1d(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int                   p0,
+            int                   p1);
+
+    // Move tensor elements by an offset given for each dimension. Elements that
+    // are shifted beyond the last position are wrapped around to the beginning.
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_roll(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int                   shift0,
+            int                   shift1,
+            int                   shift2,
+            int                   shift3);
+
+
+    // Ref: https://github.com/CompVis/stable-diffusion/blob/main/ldm/modules/diffusionmodules/util.py#L151
+    // timesteps: [N,]
+    // return: [N, dim]
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_timestep_embedding(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * timesteps,
+            int                   dim,
+            int                   max_period);
+
+    // sort rows
+    enum lm_ggml_sort_order {
+        LM_GGML_SORT_ORDER_ASC,
+        LM_GGML_SORT_ORDER_DESC,
+    };
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_argsort(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            enum lm_ggml_sort_order  order);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_arange(
+            struct lm_ggml_context * ctx,
+            float                 start,
+            float                 stop,
+            float                 step);
+
+    // top k elements per row
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_top_k(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int                   k);
+
+#define LM_GGML_KQ_MASK_PAD 64
+
+    // q:    [n_embd_k, n_batch,     n_head,    ne3 ]
+    // k:    [n_embd_k, n_kv,        n_head_kv, ne3 ]
+    // v:    [n_embd_v, n_kv,        n_head_kv, ne3 ] !! not transposed !!
+    // mask: [n_kv,     n_batch_pad, ne32,      ne33] !! n_batch_pad = LM_GGML_PAD(n_batch, LM_GGML_KQ_MASK_PAD) !!
+    // res:  [n_embd_v, n_head,      n_batch,   ne3 ] !! permuted !!
+    //
+    // broadcast:
+    //   n_head % n_head_kv == 0
+    //   n_head % ne32      == 0
+    //   ne3    % ne33      == 0
+    //
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_flash_attn_ext(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * q,
+            struct lm_ggml_tensor  * k,
+            struct lm_ggml_tensor  * v,
+            struct lm_ggml_tensor  * mask,
+            float                 scale,
+            float                 max_bias,
+            float                 logit_softcap);
+
+    LM_GGML_API void lm_ggml_flash_attn_ext_set_prec(
+            struct lm_ggml_tensor * a,
+            enum lm_ggml_prec       prec);
+
+    LM_GGML_API enum lm_ggml_prec lm_ggml_flash_attn_ext_get_prec(
+            const struct lm_ggml_tensor * a);
+
+    LM_GGML_API void lm_ggml_flash_attn_ext_add_sinks(
+            struct lm_ggml_tensor * a,
+            struct lm_ggml_tensor * sinks);
+
+    // TODO: needs to be adapted to lm_ggml_flash_attn_ext
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_flash_attn_back(
+           struct lm_ggml_context * ctx,
+           struct lm_ggml_tensor  * q,
+           struct lm_ggml_tensor  * k,
+           struct lm_ggml_tensor  * v,
+           struct lm_ggml_tensor  * d,
+           bool                  masked);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_ssm_conv(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * sx,
+            struct lm_ggml_tensor  * c);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_ssm_scan(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * s,
+            struct lm_ggml_tensor  * x,
+            struct lm_ggml_tensor  * dt,
+            struct lm_ggml_tensor  * A,
+            struct lm_ggml_tensor  * B,
+            struct lm_ggml_tensor  * C,
+            struct lm_ggml_tensor  * ids);
+
+    // partition into non-overlapping windows with padding if needed
+    // example:
+    // a:   768   64   64    1
+    // w:    14
+    // res: 768   14   14    25
+    // used in sam
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_win_part(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int                   w);
+
+    // reverse of lm_ggml_win_part
+    // used in sam
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_win_unpart(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int                   w0,
+            int                   h0,
+            int                   w);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_unary(
+            struct lm_ggml_context * ctx,
+             struct lm_ggml_tensor * a,
+             enum lm_ggml_unary_op op);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_unary_inplace(
+        struct lm_ggml_context * ctx,
+        struct lm_ggml_tensor  * a,
+        enum lm_ggml_unary_op op);
+
+    // used in sam
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_get_rel_pos(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            int                   qh,
+            int                   kh);
+
+    // used in sam
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_add_rel_pos(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * pw,
+            struct lm_ggml_tensor  * ph);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_add_rel_pos_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * pw,
+            struct lm_ggml_tensor  * ph);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_rwkv_wkv6(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * k,
+            struct lm_ggml_tensor  * v,
+            struct lm_ggml_tensor  * r,
+            struct lm_ggml_tensor  * tf,
+            struct lm_ggml_tensor  * td,
+            struct lm_ggml_tensor  * state);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_gated_linear_attn(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * k,
+            struct lm_ggml_tensor  * v,
+            struct lm_ggml_tensor  * q,
+            struct lm_ggml_tensor  * g,
+            struct lm_ggml_tensor  * state,
+            float scale);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_rwkv_wkv7(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * r,
+            struct lm_ggml_tensor  * w,
+            struct lm_ggml_tensor  * k,
+            struct lm_ggml_tensor  * v,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * b,
+            struct lm_ggml_tensor  * state);
+
+    // custom operators
+
+    typedef void (*lm_ggml_custom1_op_t)(struct lm_ggml_tensor * dst , const struct lm_ggml_tensor * a, int ith, int nth, void * userdata);
+    typedef void (*lm_ggml_custom2_op_t)(struct lm_ggml_tensor * dst , const struct lm_ggml_tensor * a, const struct lm_ggml_tensor * b, int ith, int nth, void * userdata);
+    typedef void (*lm_ggml_custom3_op_t)(struct lm_ggml_tensor * dst , const struct lm_ggml_tensor * a, const struct lm_ggml_tensor * b, const struct lm_ggml_tensor * c, int ith, int nth, void * userdata);
+
+#define LM_GGML_N_TASKS_MAX (-1)
+    // n_tasks == LM_GGML_N_TASKS_MAX means to use max number of tasks
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_map_custom1(
+            struct lm_ggml_context   * ctx,
+            struct lm_ggml_tensor    * a,
+            lm_ggml_custom1_op_t       fun,
+            int                     n_tasks,
+            void                  * userdata);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_map_custom1_inplace(
+            struct lm_ggml_context   * ctx,
+            struct lm_ggml_tensor    * a,
+            lm_ggml_custom1_op_t       fun,
+            int                     n_tasks,
+            void                  * userdata);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_map_custom2(
+            struct lm_ggml_context   * ctx,
+            struct lm_ggml_tensor    * a,
+            struct lm_ggml_tensor    * b,
+            lm_ggml_custom2_op_t       fun,
+            int                     n_tasks,
+            void                  * userdata);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_map_custom2_inplace(
+            struct lm_ggml_context   * ctx,
+            struct lm_ggml_tensor    * a,
+            struct lm_ggml_tensor    * b,
+            lm_ggml_custom2_op_t       fun,
+            int                     n_tasks,
+            void                  * userdata);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_map_custom3(
+            struct lm_ggml_context   * ctx,
+            struct lm_ggml_tensor    * a,
+            struct lm_ggml_tensor    * b,
+            struct lm_ggml_tensor    * c,
+            lm_ggml_custom3_op_t       fun,
+            int                     n_tasks,
+            void                  * userdata);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_map_custom3_inplace(
+            struct lm_ggml_context   * ctx,
+            struct lm_ggml_tensor    * a,
+            struct lm_ggml_tensor    * b,
+            struct lm_ggml_tensor    * c,
+            lm_ggml_custom3_op_t       fun,
+            int                     n_tasks,
+            void                  * userdata);
+
+    typedef void (*lm_ggml_custom_op_t)(struct lm_ggml_tensor * dst , int ith, int nth, void * userdata);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_custom_4d(
+            struct lm_ggml_context * ctx,
+            enum lm_ggml_type        type,
+            int64_t               ne0,
+            int64_t               ne1,
+            int64_t               ne2,
+            int64_t               ne3,
+            struct lm_ggml_tensor ** args,
+            int                   n_args,
+            lm_ggml_custom_op_t      fun,
+            int                   n_tasks,
+            void                * userdata);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_custom_inplace(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor ** args,
+            int                   n_args,
+            lm_ggml_custom_op_t      fun,
+            int                   n_tasks,
+            void                * userdata);
+
+    // loss function
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_cross_entropy_loss(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,  // logits
+            struct lm_ggml_tensor  * b); // labels
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_cross_entropy_loss_back(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,  // logits
+            struct lm_ggml_tensor  * b,  // labels
+            struct lm_ggml_tensor  * c); // gradients of cross_entropy_loss result
+
+    // AdamW optimizer step
+    // Paper: https://arxiv.org/pdf/1711.05101v3.pdf
+    // PyTorch: https://pytorch.org/docs/stable/generated/torch.optim.AdamW.html
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_opt_step_adamw(
+            struct lm_ggml_context * ctx,
+            struct lm_ggml_tensor  * a,
+            struct lm_ggml_tensor  * grad,
+            struct lm_ggml_tensor  * m,
+            struct lm_ggml_tensor  * v,
+            struct lm_ggml_tensor  * adamw_params); // parameters such as the learning rate
+
+    // stochastic gradient descent step (with weight decay)
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_opt_step_sgd(
+        struct lm_ggml_context * ctx,
+        struct lm_ggml_tensor *  a,
+        struct lm_ggml_tensor *  grad,
+        struct lm_ggml_tensor *  sgd_params); // alpha, weight decay
+
+    //
+    // automatic differentiation
+    //
+
+    LM_GGML_API void lm_ggml_build_forward_expand(struct lm_ggml_cgraph * cgraph, struct lm_ggml_tensor * tensor);
+    LM_GGML_API void lm_ggml_build_backward_expand(
+        struct lm_ggml_context *  ctx,        // context for gradient computation
+        struct lm_ggml_cgraph  *  cgraph,
+        struct lm_ggml_tensor  ** grad_accs);
+
+    // graph allocation in a context
+    LM_GGML_API struct lm_ggml_cgraph * lm_ggml_new_graph       (struct lm_ggml_context * ctx); // size = LM_GGML_DEFAULT_GRAPH_SIZE, grads = false
+    LM_GGML_API struct lm_ggml_cgraph * lm_ggml_new_graph_custom(struct lm_ggml_context * ctx, size_t size, bool grads);
+    LM_GGML_API struct lm_ggml_cgraph * lm_ggml_graph_dup       (struct lm_ggml_context * ctx, struct lm_ggml_cgraph * cgraph, bool force_grads);
+    LM_GGML_API void                 lm_ggml_graph_cpy       (struct lm_ggml_cgraph * src, struct lm_ggml_cgraph * dst);
+    LM_GGML_API void                 lm_ggml_graph_reset     (struct lm_ggml_cgraph * cgraph); // set regular grads + optimizer momenta to 0, set loss grad to 1
+    LM_GGML_API void                 lm_ggml_graph_clear     (struct lm_ggml_cgraph * cgraph);
+
+    LM_GGML_API int                   lm_ggml_graph_size   (struct lm_ggml_cgraph * cgraph);
+    LM_GGML_API struct lm_ggml_tensor *  lm_ggml_graph_node   (struct lm_ggml_cgraph * cgraph, int i); // if i < 0, returns nodes[n_nodes + i]
+    LM_GGML_API struct lm_ggml_tensor ** lm_ggml_graph_nodes  (struct lm_ggml_cgraph * cgraph);
+    LM_GGML_API int                   lm_ggml_graph_n_nodes(struct lm_ggml_cgraph * cgraph);
+
+    LM_GGML_API void   lm_ggml_graph_add_node(struct lm_ggml_cgraph * cgraph, struct lm_ggml_tensor * tensor);
+
+    LM_GGML_API size_t lm_ggml_graph_overhead(void);
+    LM_GGML_API size_t lm_ggml_graph_overhead_custom(size_t size, bool grads);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_graph_get_tensor  (const struct lm_ggml_cgraph * cgraph, const char * name);
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_graph_get_grad    (const struct lm_ggml_cgraph * cgraph, const struct lm_ggml_tensor * node);
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_graph_get_grad_acc(const struct lm_ggml_cgraph * cgraph, const struct lm_ggml_tensor * node);
+
+    // print info and performance information for the graph
+    LM_GGML_API void lm_ggml_graph_print(const struct lm_ggml_cgraph * cgraph);
+
+    // dump the graph into a file using the dot format
+    LM_GGML_API void lm_ggml_graph_dump_dot(const struct lm_ggml_cgraph * gb, const struct lm_ggml_cgraph * gf, const char * filename);
+
+    // TODO these functions were sandwiched in the old optimization interface, is there a better place for them?
+    typedef void (*lm_ggml_log_callback)(enum lm_ggml_log_level level, const char * text, void * user_data);
+
+    // Set callback for all future logging events.
+    // If this is not called, or NULL is supplied, everything is output on stderr.
+    LM_GGML_API void lm_ggml_log_set(lm_ggml_log_callback log_callback, void * user_data);
+
+    LM_GGML_API struct lm_ggml_tensor * lm_ggml_set_zero(struct lm_ggml_tensor * tensor);
+
+    //
+    // quantization
+    //
+
+    // - lm_ggml_quantize_init can be called multiple times with the same type
+    //   it will only initialize the quantization tables for the first call or after lm_ggml_quantize_free
+    //   automatically called by lm_ggml_quantize_chunk for convenience
+    //
+    // - lm_ggml_quantize_free will free any memory allocated by lm_ggml_quantize_init
+    //   call this at the end of the program to avoid memory leaks
+    //
+    // note: these are thread-safe
+    //
+    LM_GGML_API void lm_ggml_quantize_init(enum lm_ggml_type type);
+    LM_GGML_API void lm_ggml_quantize_free(void);
+
+    // some quantization type cannot be used without an importance matrix
+    LM_GGML_API bool lm_ggml_quantize_requires_imatrix(enum lm_ggml_type type);
+
+    // calls lm_ggml_quantize_init internally (i.e. can allocate memory)
+    LM_GGML_API size_t lm_ggml_quantize_chunk(
+            enum lm_ggml_type   type,
+               const float * src,
+                      void * dst,
+                   int64_t   start,
+                   int64_t   nrows,
+                   int64_t   n_per_row,
+               const float * imatrix);
+
+#ifdef __cplusplus
+    // restrict not standard in C++
+#    if defined(__GNUC__)
+#        define LM_GGML_RESTRICT __restrict__
+#    elif defined(__clang__)
+#        define LM_GGML_RESTRICT __restrict
+#    elif defined(_MSC_VER)
+#        define LM_GGML_RESTRICT __restrict
+#    else
+#        define LM_GGML_RESTRICT
+#    endif
+#else
+#    if defined (_MSC_VER) && (__STDC_VERSION__ < 201112L)
+#        define LM_GGML_RESTRICT __restrict
+#    else
+#        define LM_GGML_RESTRICT restrict
+#    endif
+#endif
+    typedef void (*lm_ggml_to_float_t)  (const void  * LM_GGML_RESTRICT x, float * LM_GGML_RESTRICT y, int64_t k);
+    typedef void (*lm_ggml_from_float_t)(const float * LM_GGML_RESTRICT x, void  * LM_GGML_RESTRICT y, int64_t k);
+
+    struct lm_ggml_type_traits {
+        const char             * type_name;
+        int64_t                  blck_size;
+        int64_t                  blck_size_interleave; // interleave elements in blocks
+        size_t                   type_size;
+        bool                     is_quantized;
+        lm_ggml_to_float_t          to_float;
+        lm_ggml_from_float_t        from_float_ref;
+    };
+
+    LM_GGML_API const struct lm_ggml_type_traits * lm_ggml_get_type_traits(enum lm_ggml_type type);
+
+    // ggml threadpool
+    // TODO: currently, only a few functions are in the base ggml API, while the rest are in the CPU backend
+    // the goal should be to create an API that other backends can use move everything to the ggml base
+
+    // scheduling priorities
+    enum lm_ggml_sched_priority {
+        LM_GGML_SCHED_PRIO_LOW = -1,
+        LM_GGML_SCHED_PRIO_NORMAL,
+        LM_GGML_SCHED_PRIO_MEDIUM,
+        LM_GGML_SCHED_PRIO_HIGH,
+        LM_GGML_SCHED_PRIO_REALTIME
+    };
+
+    // threadpool params
+    // Use lm_ggml_threadpool_params_default() or lm_ggml_threadpool_params_init() to populate the defaults
+    struct lm_ggml_threadpool_params {
+        bool                cpumask[LM_GGML_MAX_N_THREADS]; // mask of cpu cores (all-zeros means use default affinity settings)
+        int                 n_threads;                   // number of threads
+        enum lm_ggml_sched_priority prio;                   // thread priority
+        uint32_t            poll;                        // polling level (0 - no polling, 100 - aggressive polling)
+        bool                strict_cpu;                  // strict cpu placement
+        bool                paused;                      // start in paused state
+    };
+
+    struct lm_ggml_threadpool;     // forward declaration, see ggml.c
+
+    typedef struct lm_ggml_threadpool * lm_ggml_threadpool_t;
+
+    LM_GGML_API struct lm_ggml_threadpool_params lm_ggml_threadpool_params_default(int n_threads);
+    LM_GGML_API void                          lm_ggml_threadpool_params_init   (struct lm_ggml_threadpool_params * p, int n_threads);
+    LM_GGML_API bool                          lm_ggml_threadpool_params_match  (const struct lm_ggml_threadpool_params * p0, const struct lm_ggml_threadpool_params * p1);
+
+#ifdef  __cplusplus
+}
+#endif