Skip to content

Commit

Permalink
wip : rpi4 support
Browse files Browse the repository at this point in the history
  • Loading branch information
@ggerganov
ggerganov committed Oct 5, 2022
1 parent ce1fe95 commit 1673245
Show file tree
Hide file tree
Showing 4 changed files with 151 additions and 21 deletions.
23 changes: 16 additions & 7 deletions Makefile
View file
Original file line number Diff line number Diff line change
Expand Up @@ -15,25 +15,34 @@ CXXFLAGS += -Wall -Wextra -Wno-unused-parameter -Wno-unused-function
# OS specific
# TODO: support Windows
ifeq ($(UNAME_S),Linux)
CFLAGS += -pthread
CFLAGS += -pthread
CXXFLAGS += -pthread
endif
ifeq ($(UNAME_S),Darwin)
CFLAGS += -pthread
CFLAGS += -pthread
CXXFLAGS += -pthread
endif

# Architecture specific
ifeq ($(UNAME_P),x86_64)
CFLAGS += -mavx -mavx2 -mfma -mf16c
endif
ifneq ($(filter arm%,$(UNAME_P)),)
CFLAGS += -mfpu=neon
# Mac M1
endif
ifneq ($(filter aarch64%,$(UNAME_M)),)
CFLAGS += -mfpu=neon
ifneq ($(filter aarch64%,$(UNAME_P)),)
endif
ifneq ($(filter armv6%,$(UNAME_M)),)
# Raspberry Pi 1, 2, 3
CFLAGS += -mfpu=neon-fp-armv8 -mfp16-format=ieee -mno-unaligned-access
endif
ifneq ($(filter armv%,$(UNAME_M)),)
ifneq ($(filter armv7%,$(UNAME_M)),)
# Raspberry Pi 4
CFLAGS += -mcpu=cortex-a72 -mfloat-abi=hard -mfpu=neon-fp-armv8 -mfp16-format=ieee -mno-unaligned-access
CFLAGS += -mfpu=neon-fp-armv8 -mfp16-format=ieee -mno-unaligned-access -funsafe-math-optimizations

This comment has been minimized.

Sign in to view

Copy link
@clach04

clach04 Feb 5, 2023

Contributor

Comment mentions Raspberry Pi 4 - but the filter is for 32-bit and the mfpu is set to 64-bit.

Can you clarify this part of the makefile? Is this supposed to be for 32-bit OS for Raspberry Pi 4? I suspect expect it to be using 32-bit flags. Cross reference to #470
endif
ifneq ($(filter armv8%,$(UNAME_M)),)
# Raspberry Pi 4
CFLAGS += -mfp16-format=ieee -mno-unaligned-access
endif

#
Expand Down
130 changes: 123 additions & 7 deletions ggml.c
View file
Original file line number Diff line number Diff line change
@@ -1,5 +1,6 @@
#include "ggml.h"

#include <alloca.h>
#include <assert.h>
#include <time.h>
#include <math.h>
Expand All @@ -12,7 +13,12 @@
#include <pthread.h>

#define GGML_DEBUG 0
#define GGML_MEM_ALIGN 16

#if UINTPTR_MAX == 0xFFFFFFFF
#define GGML_MEM_ALIGN 4
#else
#define GGML_MEM_ALIGN 16
#endif

#define MAX(a, b) ((a) > (b) ? (a) : (b))
#define MIN(a, b) ((a) < (b) ? (a) : (b))
Expand Down Expand Up @@ -305,6 +311,7 @@ inline static void ggml_vec_dot_f16(const int n, float * restrict s, ggml_fp16_t
#ifdef __ARM_NEON
const int n32 = (n & ~31);

#if defined(__ARM_FEATURE_FP16_VECTOR_ARITHMETIC)
float16x8_t sum0 = vdupq_n_f16(0);
float16x8_t sum1 = vdupq_n_f16(0);
float16x8_t sum2 = vdupq_n_f16(0);
Expand Down Expand Up @@ -344,6 +351,61 @@ inline static void ggml_vec_dot_f16(const int n, float * restrict s, ggml_fp16_t

float32x2_t sumf32 = vadd_f32(vget_low_f32(sum0f32), vget_high_f32(sum0f32));
sumf = vget_lane_f32(sumf32, 0) + vget_lane_f32(sumf32, 1);
#else
float32x4_t sum0 = vdupq_n_f32(0);
float32x4_t sum1 = vdupq_n_f32(0);
float32x4_t sum2 = vdupq_n_f32(0);
float32x4_t sum3 = vdupq_n_f32(0);
float32x4_t sum4 = vdupq_n_f32(0);
float32x4_t sum5 = vdupq_n_f32(0);
float32x4_t sum6 = vdupq_n_f32(0);
float32x4_t sum7 = vdupq_n_f32(0);

float32x4_t x0, x1, x2, x3, x4, x5, x6, x7;
float32x4_t y0, y1, y2, y3, y4, y5, y6, y7;

for (int i = 0; i < n32; i += 32) {
x0 = vcvt_f32_f16(vld1_f16(x + i + 0 ));
x1 = vcvt_f32_f16(vld1_f16(x + i + 4 ));
x2 = vcvt_f32_f16(vld1_f16(x + i + 8 ));
x3 = vcvt_f32_f16(vld1_f16(x + i + 12));
x4 = vcvt_f32_f16(vld1_f16(x + i + 16));
x5 = vcvt_f32_f16(vld1_f16(x + i + 20));
x6 = vcvt_f32_f16(vld1_f16(x + i + 24));
x7 = vcvt_f32_f16(vld1_f16(x + i + 28));

y0 = vcvt_f32_f16(vld1_f16(y + i + 0 ));
y1 = vcvt_f32_f16(vld1_f16(y + i + 4 ));
y2 = vcvt_f32_f16(vld1_f16(y + i + 8 ));
y3 = vcvt_f32_f16(vld1_f16(y + i + 12));
y4 = vcvt_f32_f16(vld1_f16(y + i + 16));
y5 = vcvt_f32_f16(vld1_f16(y + i + 20));
y6 = vcvt_f32_f16(vld1_f16(y + i + 24));
y7 = vcvt_f32_f16(vld1_f16(y + i + 28));

sum0 = vfmaq_f32(sum0, x0, y0);
sum1 = vfmaq_f32(sum1, x1, y1);
sum2 = vfmaq_f32(sum2, x2, y2);
sum3 = vfmaq_f32(sum3, x3, y3);
sum4 = vfmaq_f32(sum4, x4, y4);
sum5 = vfmaq_f32(sum5, x5, y5);
sum6 = vfmaq_f32(sum6, x6, y6);
sum7 = vfmaq_f32(sum7, x7, y7);
}

// reduce sum0..sum7 to sum0
sum0 = vaddq_f32(sum0, sum1);
sum2 = vaddq_f32(sum2, sum3);
sum4 = vaddq_f32(sum4, sum5);
sum6 = vaddq_f32(sum6, sum7);
sum0 = vaddq_f32(sum0, sum2);
sum4 = vaddq_f32(sum4, sum6);
sum0 = vaddq_f32(sum0, sum4);

// reduce sum0 to sumf
float32x2_t sumf32 = vadd_f32(vget_low_f32(sum0), vget_high_f32(sum0));
sumf = vget_lane_f32(sumf32, 0) + vget_lane_f32(sumf32, 1);
#endif

// leftovers
for (int i = n32; i < n; ++i) {
Expand Down Expand Up @@ -486,6 +548,7 @@ inline static void ggml_vec_mad_f16(const int n, ggml_fp16_t * restrict y, ggml_
// NEON 128-bit
const int n32 = (n & ~31);

#if defined(__ARM_FEATURE_FP16_VECTOR_ARITHMETIC)
const float16x8_t v8 = vdupq_n_f16(v);

float16x8_t x0, x1, x2, x3;
Expand All @@ -512,6 +575,51 @@ inline static void ggml_vec_mad_f16(const int n, ggml_fp16_t * restrict y, ggml_
vst1q_f16(y + i + 16, y2);
vst1q_f16(y + i + 24, y3);
}
#else
const float32x4_t v40 = vdupq_n_f32(v);
const float32x4_t v41 = vdupq_n_f32(v);

float32x4_t x0, x1, x2, x3, x4, x5, x6, x7;
float32x4_t y0, y1, y2, y3, y4, y5, y6, y7;

for (int i = 0; i < n32; i += 32) {
y0 = vcvt_f32_f16(vld1_f16(y + i + 0 ));
y1 = vcvt_f32_f16(vld1_f16(y + i + 4 ));
y2 = vcvt_f32_f16(vld1_f16(y + i + 8 ));
y3 = vcvt_f32_f16(vld1_f16(y + i + 12));
y4 = vcvt_f32_f16(vld1_f16(y + i + 16));
y5 = vcvt_f32_f16(vld1_f16(y + i + 20));
y6 = vcvt_f32_f16(vld1_f16(y + i + 24));
y7 = vcvt_f32_f16(vld1_f16(y + i + 28));

x0 = vcvt_f32_f16(vld1_f16(x + i + 0 ));
x1 = vcvt_f32_f16(vld1_f16(x + i + 4 ));
x2 = vcvt_f32_f16(vld1_f16(x + i + 8 ));
x3 = vcvt_f32_f16(vld1_f16(x + i + 12));
x4 = vcvt_f32_f16(vld1_f16(x + i + 16));
x5 = vcvt_f32_f16(vld1_f16(x + i + 20));
x6 = vcvt_f32_f16(vld1_f16(x + i + 24));
x7 = vcvt_f32_f16(vld1_f16(x + i + 28));

y0 = vfmaq_f32(y0, x0, v40);
y1 = vfmaq_f32(y1, x1, v40);
y2 = vfmaq_f32(y2, x2, v40);
y3 = vfmaq_f32(y3, x3, v40);
y4 = vfmaq_f32(y4, x4, v41);
y5 = vfmaq_f32(y5, x5, v41);
y6 = vfmaq_f32(y6, x6, v41);
y7 = vfmaq_f32(y7, x7, v41);

vst1_f16(y + i + 0 , vcvt_f16_f32(y0));
vst1_f16(y + i + 4 , vcvt_f16_f32(y1));
vst1_f16(y + i + 8 , vcvt_f16_f32(y2));
vst1_f16(y + i + 12, vcvt_f16_f32(y3));
vst1_f16(y + i + 16, vcvt_f16_f32(y4));
vst1_f16(y + i + 20, vcvt_f16_f32(y5));
vst1_f16(y + i + 24, vcvt_f16_f32(y6));
vst1_f16(y + i + 28, vcvt_f16_f32(y7));
}
#endif

// leftovers
for (int i = n32; i < n; ++i) {
Expand Down Expand Up @@ -911,16 +1019,18 @@ struct ggml_context * ggml_init(struct ggml_init_params params) {
if (is_first_call) {
const uint64_t t_start = ggml_time_us(); UNUSED(t_start);

ggml_fp16_t ii;
for (int i = 0; i < (1 << 16); ++i) {
uint16_t ii = (uint16_t) i;
const float f = ggml_fp16_to_fp32(*(ggml_fp16_t *)(&ii));
uint16_t ui = i;
memcpy(&ii, &ui, sizeof(ii));
const float f = ggml_fp16_to_fp32(ii);
table_gelu_f16[i] = ggml_fp32_to_fp16(ggml_gelu_f32(f));
table_exp_f16[i] = ggml_fp32_to_fp16(exp(f));
}

const uint64_t t_end = ggml_time_us(); UNUSED(t_end);

GGML_PRINT_DEBUG("%s: GELU table initialized in %f ms\n", __func__, (t_end - t_start)/1000.0f);
GGML_PRINT_DEBUG("%s: GELU and EXP tables initialized in %f ms\n", __func__, (t_end - t_start)/1000.0f);

is_first_call = false;
}
Expand Down Expand Up @@ -4427,13 +4537,15 @@ void ggml_compute_forward_soft_max_f32(

ggml_float sum = 0.0;

uint16_t ss;
for (int i = 0; i < nc; i++) {
if (p[i] == -INFINITY) {
p[i] = 0.0;
} else {
//const float val = (p[i] == -INFINITY) ? 0.0 : exp(p[i] - max);
ggml_fp16_t s = ggml_fp32_to_fp16(p[i] - max);
const float val = ggml_fp16_to_fp32(table_exp_f16[*(uint16_t *) &s]);
memcpy(&ss, &s, sizeof(ss));
const float val = ggml_fp16_to_fp32(table_exp_f16[ss]);
sum += val;
p[i] = val;
}
Expand Down Expand Up @@ -5234,13 +5346,15 @@ void ggml_compute_forward_flash_attn_f32(

ggml_float sum = 0.0;

uint16_t ss;
for (int i = 0; i < M; i++) {
if (S[i] == -INFINITY) {
S[i] = 0.0;
} else {
//const float val = (S[i] == -INFINITY) ? 0.0 : exp(S[i] - max);
ggml_fp16_t s = ggml_fp32_to_fp16(S[i] - max);
const float val = ggml_fp16_to_fp32(table_exp_f16[*(uint16_t *) &s]);
memcpy(&ss, &s, sizeof(ss));
const float val = ggml_fp16_to_fp32(table_exp_f16[ss]);
sum += val;
S[i] = val;
}
Expand Down Expand Up @@ -5413,13 +5527,15 @@ void ggml_compute_forward_flash_attn_f16(

ggml_float sum = 0.0;

uint16_t ss;
for (int i = 0; i < M; i++) {
if (S[i] == -INFINITY) {
S[i] = 0.0;
} else {
//const float val = (S[i] == -INFINITY) ? 0.0 : exp(S[i] - max);
ggml_fp16_t s = ggml_fp32_to_fp16(S[i] - max);
const float val = ggml_fp16_to_fp32(table_exp_f16[*(uint16_t *) &s]);
memcpy(&ss, &s, sizeof(ss));
const float val = ggml_fp16_to_fp32(table_exp_f16[ss]);
sum += val;
S[i] = val;
}
Expand Down
2 changes: 1 addition & 1 deletion ggml.h
View file
Original file line number Diff line number Diff line change
Expand Up @@ -108,7 +108,7 @@ struct ggml_tensor {
int64_t perf_time_us;

void * data;
char pad[8];
char padding[8];
};

// computation graph
Expand Down
17 changes: 11 additions & 6 deletions whisper.cpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -1291,7 +1291,8 @@ bool whisper_encode(
struct ggml_tensor * inpO = ggml_add(ctxL, cur, inpFF);

{
struct ggml_cgraph gf = { .n_threads = n_threads };
struct ggml_cgraph gf = {};
gf.n_threads = n_threads;

ggml_build_forward_expand(&gf, inpO);
ggml_graph_compute (ctxL, &gf);
Expand Down Expand Up @@ -1327,7 +1328,8 @@ bool whisper_encode(

// run the computation
{
struct ggml_cgraph gf = { .n_threads = n_threads };
struct ggml_cgraph gf = {};
gf.n_threads = n_threads;

ggml_build_forward_expand(&gf, cur);
ggml_graph_compute (ctx0, &gf);
Expand All @@ -1351,7 +1353,8 @@ bool whisper_encode(

// pre-compute cross-attention memory
{
struct ggml_cgraph gf = { .n_threads = n_threads };
struct ggml_cgraph gf = {};
gf.n_threads = n_threads;

// TODO: hack to disconnect the encoded features from the previous graph
cur->op = GGML_OP_NONE;
Expand Down Expand Up @@ -1461,7 +1464,8 @@ bool whisper_decode(
};

struct ggml_context * ctxL = ggml_init(paramsL);
struct ggml_cgraph gf = { .n_threads = n_threads };
struct ggml_cgraph gf = {};
gf.n_threads = n_threads;

// norm
{
Expand Down Expand Up @@ -1744,7 +1748,8 @@ bool whisper_decode(

// run the computation
{
struct ggml_cgraph gf = { .n_threads = n_threads };
struct ggml_cgraph gf = {};
gf.n_threads = n_threads;

ggml_build_forward_expand(&gf, cur);
ggml_graph_compute (ctx0, &gf);
Expand Down Expand Up @@ -2334,7 +2339,7 @@ int whisper_full(
}
}

if (seek >= whisper_n_len(ctx)) {
if (seek + 100 >= whisper_n_len(ctx)) {
break;
}

Expand Down

0 comments on commit 1673245

Please sign in to comment.