vrr/koboldcpp
2
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2025-09-10 17:14:36 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

fixed compiler warnings

Browse source
This commit is contained in:
Concedo 2024-04-08 16:40:55 +08:00
parent 909e4334f9
commit 125f84aa02
3 changed files with 17 additions and 17 deletions
Download patch file Download diff file Expand all files Collapse all files

2
gpttype_adapter.cpp
View file

@ -2313,7 +2313,7 @@ generation_outputs gpttype_generate(const generation_inputs inputs)
float pt2 = (time2*1000.0/(realnpredict==0?1:realnpredict));
float ts2 = (1000.0/pt2);
float tokens_per_second = (realnpredict == 0 ? 0 : realnpredict / (time1 + time2));
printf("\nCtxLimit: %d/%d, Process:%.2fs (%.1fms/T = %.2fT/s), Generate:%.2fs (%.1fms/T = %.2fT/s), Total:%.2fs (%.2fT/s)",current_context_tokens.size(),nctx, time1, pt1, ts1, time2, pt2, ts2, (time1 + time2), tokens_per_second);
printf("\nCtxLimit: %d/%d, Process:%.2fs (%.1fms/T = %.2fT/s), Generate:%.2fs (%.1fms/T = %.2fT/s), Total:%.2fs (%.2fT/s)",(int)current_context_tokens.size(),(int)nctx, time1, pt1, ts1, time2, pt2, ts2, (time1 + time2), tokens_per_second);
fflush(stdout);
output.status = 1;
generation_finished = true;

24
otherarch/ggml_v3-cuda.cu
View file

@ -627,18 +627,18 @@ static __device__ __forceinline__ float warp_reduce_max(float x) {
return x;
}
static __device__ __forceinline__ half2 warp_reduce_max(half2 x) {
#if !(defined(GGML_USE_HIPBLAS) && defined(__HIP_PLATFORM_AMD__)) && __CUDA_ARCH__ >= CC_PASCAL && CUDART_VERSION >= CUDART_HMAX
#pragma unroll
for (int mask = 16; mask > 0; mask >>= 1) {
x = __hmax2(x, __shfl_xor_sync(0xffffffff, x, mask, 32));
}
return x;
#else
(void) x;
bad_arch();
#endif // !(defined(GGML_USE_HIPBLAS) && defined(__HIP_PLATFORM_AMD__)) && __CUDA_ARCH__ >= CC_PASCAL && CUDART_VERSION >= CUDART_HMAX
}
// static __device__ __forceinline__ half2 warp_reduce_max(half2 x) {
// #if !(defined(GGML_USE_HIPBLAS) && defined(__HIP_PLATFORM_AMD__)) && __CUDA_ARCH__ >= CC_PASCAL && CUDART_VERSION >= CUDART_HMAX
// #pragma unroll
// for (int mask = 16; mask > 0; mask >>= 1) {
// x = __hmax2(x, __shfl_xor_sync(0xffffffff, x, mask, 32));
// }
// return x;
// #else
// (void) x;
// bad_arch();
// #endif // !(defined(GGML_USE_HIPBLAS) && defined(__HIP_PLATFORM_AMD__)) && __CUDA_ARCH__ >= CC_PASCAL && CUDART_VERSION >= CUDART_HMAX
// }
static __device__ __forceinline__ float op_repeat(const float a, const float b) {
return b;

8
otherarch/sdcpp/sdtype_adapter.cpp
View file

@ -326,7 +326,7 @@ sd_generation_outputs sdtype_generate(const sd_generation_inputs inputs)
if(!is_quiet && sddebugmode==1)
{
printf("\nTXT2IMG PROMPT:%s\nNPROMPT:%s\nCLPSKP:%d\nCFGSCLE:%f\nW:%d\nH:%d\nSM:%d\nSTEP:%d\nSEED:%d\nBATCH:%d\nCIMG:%d\nCSTR:%f\n\n",
printf("\nTXT2IMG PROMPT:%s\nNPROMPT:%s\nCLPSKP:%d\nCFGSCLE:%f\nW:%d\nH:%d\nSM:%d\nSTEP:%d\nSEED:%d\nBATCH:%d\nCIMG:%p\nCSTR:%f\n\n",
sd_params->prompt.c_str(),
sd_params->negative_prompt.c_str(),
sd_params->clip_skip,
@ -335,7 +335,7 @@ sd_generation_outputs sdtype_generate(const sd_generation_inputs inputs)
sd_params->height,
sd_params->sample_method,
sd_params->sample_steps,
sd_params->seed,
(int)sd_params->seed,
sd_params->batch_count,
control_image,
sd_params->control_strength);
@ -392,7 +392,7 @@ sd_generation_outputs sdtype_generate(const sd_generation_inputs inputs)
if(!is_quiet && sddebugmode==1)
{
printf("\nIMG2IMG PROMPT:%s\nNPROMPT:%s\nCLPSKP:%d\nCFGSCLE:%f\nW:%d\nH:%d\nSM:%d\nSTEP:%d\nSEED:%d\nBATCH:%d\nCIMG:%d\nSTR:%f\n\n",
printf("\nIMG2IMG PROMPT:%s\nNPROMPT:%s\nCLPSKP:%d\nCFGSCLE:%f\nW:%d\nH:%d\nSM:%d\nSTEP:%d\nSEED:%d\nBATCH:%d\nCIMG:%p\nSTR:%f\n\n",
sd_params->prompt.c_str(),
sd_params->negative_prompt.c_str(),
sd_params->clip_skip,
@ -401,7 +401,7 @@ sd_generation_outputs sdtype_generate(const sd_generation_inputs inputs)
sd_params->height,
sd_params->sample_method,
sd_params->sample_steps,
sd_params->seed,
(int)sd_params->seed,
sd_params->batch_count,
control_image,
sd_params->strength);