matt/ollama37
1
0
Fork 0
mirror of https://github.com/dogkeeper886/ollama37.git synced 2025-12-10 15:57:04 +00:00
Code Issues Packages Projects Releases Wiki Activity

Report more information about GPUs in verbose mode

Browse Source 
This adds additional calls to both CUDA and ROCm management libraries to
discover additional attributes about the GPU(s) detected in the system, and
wires up runtime verbosity selection.  When users hit problems with GPUs we can
ask them to run with `OLLAMA_DEBUG=1 ollama serve` and share the results.
This commit is contained in:
Daniel Hiltgen
2024-01-22 16:03:32 -08:00
parent 0759d8996e
commit 987c16b2f7
6 changed files with 171 additions and 30 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
gpu/gpu_info_cuda.h
View File

@@ -15,14 +15,26 @@ typedef struct nvmlMemory_st {
unsigned long long used;
} nvmlMemory_t;
typedef enum nvmlBrandType_enum
{
NVML_BRAND_UNKNOWN = 0,
} nvmlBrandType_t;
typedef struct cuda_handle {
void *handle;
uint16_t verbose;
nvmlReturn_t (*initFn)(void);
nvmlReturn_t (*shutdownFn)(void);
nvmlReturn_t (*getHandle)(unsigned int, nvmlDevice_t *);
nvmlReturn_t (*getMemInfo)(nvmlDevice_t, nvmlMemory_t *);
nvmlReturn_t (*getCount)(unsigned int *);
nvmlReturn_t (*getComputeCapability)(nvmlDevice_t, int* major, int* minor);
nvmlReturn_t (*nvmlSystemGetDriverVersion) (char* version, unsigned int length);
nvmlReturn_t (*nvmlDeviceGetName) (nvmlDevice_t device, char* name, unsigned int length);
nvmlReturn_t (*nvmlDeviceGetSerial) (nvmlDevice_t device, char* serial, unsigned int length);
nvmlReturn_t (*nvmlDeviceGetVbiosVersion) (nvmlDevice_t device, char* version, unsigned int length);
nvmlReturn_t (*nvmlDeviceGetBoardPartNumber) (nvmlDevice_t device, char* partNumber, unsigned int length);
nvmlReturn_t (*nvmlDeviceGetBrand) (nvmlDevice_t device, nvmlBrandType_t* type);
} cuda_handle_t;
typedef struct cuda_init_resp {