I have built with DGGML_CUDA=ON via cmake but does not use gpu

[TeaCult]

I have built as instructed, my cuda libs are working in text-generaton-webui and torch but its not working with llama-cli or llama-server :

gediz@LLAMA ~/llama.cpp $ ./llama-server -np 4 -c 8192 -n 512 -ngl 100 -m ../text-generation-webui/models/gemma-2-27b-it-SimPO-37K-Q4_K_M.gguf
warning: not compiled with GPU offload support, --gpu-layers option will be ignored
warning: see main README.md for information on enabling GPU BLAS support
build: 3891 (d5cb8684) with cc (GCC) 13.1.1 20230714 for x86_64-pc-linux-gnu
system info: n_threads = 16, n_threads_batch = 16, total_threads = 16

system_info: n_threads = 16 (n_threads_batch = 16) / 16 | AVX = 1 | AVX_VNNI = 0 | AVX2 = 1 | AVX512 = 0 | AVX512_VBMI = 0 | AVX512_VNNI = 0 | AVX512_BF16 = 0 | FMA = 1 | NEON = 0 | SVE = 0 | ARM_FMA = 0 | F16C = 1 | FP16_VA = 0 | RISCV_VECT = 0 | WASM_SIMD = 0 | BLAS = 0 | SSE3 = 1 | SSSE3 = 1 | VSX = 0 | MATMUL_INT8 = 0 | LLAMAFILE = 1 | 

main: HTTP server is listening, hostname: 127.0.0.1, port: 8080, http threads: 15
main: loading model
llama_model_loader: loaded meta data with 44 key-value pairs and 508 tensors from ../text-generation-webui/models/gemma-2-27b-it-SimPO-37K-Q4_K_M.gguf (version GGUF V3 (latest))
llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
llama_model_loader: - kv   0:                       general.architecture str              = gemma2
llama_model_loader: - kv   1:                               general.type str              = model
llama_model_loader: - kv   2:                               general.name str              = Gemma 2 27b It
llama_model_loader: - kv   3:                       general.organization str              = Google
llama_model_loader: - kv   4:                           general.finetune str              = it
llama_model_loader: - kv   5:                           general.basename str              = gemma-2
llama_model_loader: - kv   6:                         general.size_label str              = 27B
llama_model_loader: - kv   7:                            general.license str              = gemma
llama_model_loader: - kv   8:                   general.base_model.count u32              = 1
llama_model_loader: - kv   9:                  general.base_model.0.name str              = Gemma 2 27b It
llama_model_loader: - kv  10:          general.base_model.0.organization str              = Google
llama_model_loader: - kv  11:              general.base_model.0.repo_url str              = https://huggingface.co/google/gemma-2...
llama_model_loader: - kv  12:                               general.tags arr[str,3]       = ["alignment-handbook", "generated_fro...
llama_model_loader: - kv  13:                      gemma2.context_length u32              = 8192
llama_model_loader: - kv  14:                    gemma2.embedding_length u32              = 4608
llama_model_loader: - kv  15:                         gemma2.block_count u32              = 46
llama_model_loader: - kv  16:                 gemma2.feed_forward_length u32              = 36864
llama_model_loader: - kv  17:                gemma2.attention.head_count u32              = 32
llama_model_loader: - kv  18:             gemma2.attention.head_count_kv u32              = 16
llama_model_loader: - kv  19:    gemma2.attention.layer_norm_rms_epsilon f32              = 0.000001
llama_model_loader: - kv  20:                gemma2.attention.key_length u32              = 128
llama_model_loader: - kv  21:              gemma2.attention.value_length u32              = 128
llama_model_loader: - kv  22:                          general.file_type u32              = 15
llama_model_loader: - kv  23:              gemma2.attn_logit_softcapping f32              = 50.000000
llama_model_loader: - kv  24:             gemma2.final_logit_softcapping f32              = 30.000000
llama_model_loader: - kv  25:            gemma2.attention.sliding_window u32              = 4096
llama_model_loader: - kv  26:                       tokenizer.ggml.model str              = llama
llama_model_loader: - kv  27:                         tokenizer.ggml.pre str              = default
llama_model_loader: - kv  28:                      tokenizer.ggml.tokens arr[str,256000]  = ["<pad>", "<eos>", "<bos>", "<unk>", ...
llama_model_loader: - kv  29:                      tokenizer.ggml.scores arr[f32,256000]  = [-1000.000000, -1000.000000, -1000.00...
llama_model_loader: - kv  30:                  tokenizer.ggml.token_type arr[i32,256000]  = [3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, ...
llama_model_loader: - kv  31:                tokenizer.ggml.bos_token_id u32              = 2
llama_model_loader: - kv  32:                tokenizer.ggml.eos_token_id u32              = 1
llama_model_loader: - kv  33:            tokenizer.ggml.unknown_token_id u32              = 3
llama_model_loader: - kv  34:            tokenizer.ggml.padding_token_id u32              = 0
llama_model_loader: - kv  35:               tokenizer.ggml.add_bos_token bool             = true
llama_model_loader: - kv  36:               tokenizer.ggml.add_eos_token bool             = false
llama_model_loader: - kv  37:                    tokenizer.chat_template str              = {{ bos_token }}{% if messages[0]['rol...
llama_model_loader: - kv  38:            tokenizer.ggml.add_space_prefix bool             = false
llama_model_loader: - kv  39:               general.quantization_version u32              = 2
llama_model_loader: - kv  40:                      quantize.imatrix.file str              = /models_out/gemma-2-27b-it-SimPO-37K-...
llama_model_loader: - kv  41:                   quantize.imatrix.dataset str              = /training_dir/calibration_datav3.txt
llama_model_loader: - kv  42:             quantize.imatrix.entries_count i32              = 322
llama_model_loader: - kv  43:              quantize.imatrix.chunks_count i32              = 128
llama_model_loader: - type  f32:  185 tensors
llama_model_loader: - type q4_K:  278 tensors
llama_model_loader: - type q6_K:   45 tensors
llm_load_vocab: special_eos_id is not in special_eog_ids - the tokenizer config may be incorrect
llm_load_vocab: special tokens cache size = 249
llm_load_vocab: token to piece cache size = 1.6014 MB
llm_load_print_meta: format           = GGUF V3 (latest)
llm_load_print_meta: arch             = gemma2
llm_load_print_meta: vocab type       = SPM
llm_load_print_meta: n_vocab          = 256000
llm_load_print_meta: n_merges         = 0
llm_load_print_meta: vocab_only       = 0
llm_load_print_meta: n_ctx_train      = 8192
llm_load_print_meta: n_embd           = 4608
llm_load_print_meta: n_layer          = 46
llm_load_print_meta: n_head           = 32
llm_load_print_meta: n_head_kv        = 16
llm_load_print_meta: n_rot            = 128
llm_load_print_meta: n_swa            = 4096
llm_load_print_meta: n_embd_head_k    = 128
llm_load_print_meta: n_embd_head_v    = 128
llm_load_print_meta: n_gqa            = 2
llm_load_print_meta: n_embd_k_gqa     = 2048
llm_load_print_meta: n_embd_v_gqa     = 2048
llm_load_print_meta: f_norm_eps       = 0.0e+00
llm_load_print_meta: f_norm_rms_eps   = 1.0e-06
llm_load_print_meta: f_clamp_kqv      = 0.0e+00
llm_load_print_meta: f_max_alibi_bias = 0.0e+00
llm_load_print_meta: f_logit_scale    = 0.0e+00
llm_load_print_meta: n_ff             = 36864
llm_load_print_meta: n_expert         = 0
llm_load_print_meta: n_expert_used    = 0
llm_load_print_meta: causal attn      = 1
llm_load_print_meta: pooling type     = 0
llm_load_print_meta: rope type        = 2
llm_load_print_meta: rope scaling     = linear
llm_load_print_meta: freq_base_train  = 10000.0
llm_load_print_meta: freq_scale_train = 1
llm_load_print_meta: n_ctx_orig_yarn  = 8192
llm_load_print_meta: rope_finetuned   = unknown
llm_load_print_meta: ssm_d_conv       = 0
llm_load_print_meta: ssm_d_inner      = 0
llm_load_print_meta: ssm_d_state      = 0
llm_load_print_meta: ssm_dt_rank      = 0
llm_load_print_meta: ssm_dt_b_c_rms   = 0
llm_load_print_meta: model type       = 27B
llm_load_print_meta: model ftype      = Q4_K - Medium
llm_load_print_meta: model params     = 27.23 B
llm_load_print_meta: model size       = 15.50 GiB (4.89 BPW) 
llm_load_print_meta: general.name     = Gemma 2 27b It
llm_load_print_meta: BOS token        = 2 '<bos>'
llm_load_print_meta: EOS token        = 1 '<eos>'
llm_load_print_meta: UNK token        = 3 '<unk>'
llm_load_print_meta: PAD token        = 0 '<pad>'
llm_load_print_meta: LF token         = 227 '<0x0A>'
llm_load_print_meta: EOT token        = 107 '<end_of_turn>'
llm_load_print_meta: EOG token        = 1 '<eos>'
llm_load_print_meta: EOG token        = 107 '<end_of_turn>'
llm_load_print_meta: max token length = 48
llm_load_tensors: ggml ctx size =    0.23 MiB
llm_load_tensors:        CPU buffer size = 15868.49 MiB
............................................................................................
llama_new_context_with_model: n_ctx      = 8192
llama_new_context_with_model: n_batch    = 2048
llama_new_context_with_model: n_ubatch   = 512
llama_new_context_with_model: flash_attn = 0
llama_new_context_with_model: freq_base  = 10000.0
llama_new_context_with_model: freq_scale = 1
llama_kv_cache_init:        CPU KV buffer size =  2944.00 MiB
llama_new_context_with_model: KV self size  = 2944.00 MiB, K (f16): 1472.00 MiB, V (f16): 1472.00 MiB
llama_new_context_with_model:        CPU  output buffer size =     4.88 MiB
llama_new_context_with_model:        CPU compute buffer size =   578.01 MiB
llama_new_context_with_model: graph nodes  = 1850
llama_new_context_with_model: graph splits = 1
llama_init_from_gpt_params: warming up the model with an empty run - please wait ... (--no-warmup to disable)
srv          init: initializing slots, n_slots = 4
slot         init: id  0 | task -1 | new slot n_ctx_slot = 2048
slot         init: id  1 | task -1 | new slot n_ctx_slot = 2048
slot         init: id  2 | task -1 | new slot n_ctx_slot = 2048
slot         init: id  3 | task -1 | new slot n_ctx_slot = 2048
main: model loaded
main: chat template, built_in: 1, chat_example: '<start_of_turn>user
You are a helpful assistant

Hello<end_of_turn>
<start_of_turn>model
Hi there<end_of_turn>
<start_of_turn>user
How are you?<end_of_turn>
<start_of_turn>model
'
main: server is listening on 127.0.0.1:8080 - starting the main loop
srv  update_slots: all slots are idle

Could you help me out please

gediz@LLAMA ~/llama.cpp $ cmake -B build -DGGML_CUDA=ON
-- The C compiler identification is GNU 13.1.1
-- The CXX compiler identification is GNU 13.1.1
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Check for working C compiler: /usr/bin/cc - skipped
-- Detecting C compile features
-- Detecting C compile features - done
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Check for working CXX compiler: /usr/bin/c++ - skipped
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Found Git: /usr/bin/git (found version "2.41.0") 
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD - Success
-- Found Threads: TRUE  
-- Found OpenMP_C: -fopenmp (found version "4.5") 
-- Found OpenMP_CXX: -fopenmp (found version "4.5") 
-- Found OpenMP: TRUE (found version "4.5")  
-- OpenMP found
-- Using llamafile
-- Found CUDAToolkit: /opt/cuda/include (found version "12.2.91") 
-- CUDA found
-- Using CUDA architectures: 52;61;70;75
-- The CUDA compiler identification is NVIDIA 12.2.91
-- Detecting CUDA compiler ABI info
-- Detecting CUDA compiler ABI info - done
-- Check for working CUDA compiler: /opt/cuda/bin/nvcc - skipped
-- Detecting CUDA compile features
-- Detecting CUDA compile features - done
-- CUDA host compiler is GNU 12.3.0

-- Warning: ccache not found - consider installing it for faster compilation or disable this warning with GGML_CCACHE=OFF
-- CMAKE_SYSTEM_PROCESSOR: x86_64
-- x86 detected
-- Configuring done (5.3s)
-- Generating done (0.1s)
-- Build files have been written to: /home/gediz/llama.cpp/build
gediz@LLAMA ~/llama.cpp $ cmake --build build --config Release

I am on archlinux Linux LLAMA 6.4.6-arch1-1 #1 SMP PREEMPT_DYNAMIC Mon, 24 Jul 2023 20:19:38 +0000 x86_64 GNU/Linux
| NVIDIA-SMI 535.86.05 Driver Version: 535.86.05 CUDA Version: 12.2

[danbev]

It looks like you might be running the wrong executable:

warning: not compiled with GPU offload support, --gpu-layers option will be ignored

When building with CMake like you showed above I believe the executable files will be placed in build/bin.
So perhaps something like this might work:

$ build/bin/llama-server -np 4 -c 8192 -n 512 -ngl 100 -m ../text-generation-webui/models/gemma-2-27b-it-SimPO-37K-Q4_K_M.gguf

It could be that you previously built a llama-server using make which would place the executable in the current directory, and it might not have been compiled with CUDA support enabled.

[lexasub]

@TeaCult, do clean build in new folder with inital enable -DGGML_GPU=ON