Closed
Description
Name and Version
Attaching to llamacpp-1
llamacpp-1 | ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no
llamacpp-1 | ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
llamacpp-1 | ggml_cuda_init: found 1 CUDA devices:
llamacpp-1 | Device 0: NVIDIA GeForce GTX 1070, compute capability 6.1, VMM: yes
llamacpp-1 | version: 4932 (9ffcc9e)
llamacpp-1 | built with cc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0 for x86_64-linux-gnu
Operating systems
Linux
Which llama.cpp modules do you know to be affected?
llama-server, Other (Please specify in the next section), llama-cli
Command line
docker server-cuda and light-cuda
--port 80 -c 512 --verbose -c 8192 -ngl 60 -m /models/Llama-3.2-3B-Instruct-Q6_K_L.ggufProblem description & steps to reproduce
It crashes on executing a prompt, even if warmup is disabled, has soon as a prompt is inputed llamacpp crashes
it does not crash if it is not compile with BMI2
-DGGML_BMI2=OFF
This means that it is not detecting BMI2 CPU correctly on ggml/src/ggml-cpu/cpu-feats-x86.cpp
probably bool BMI2(void) { return f_7_ebx[8]; } or data from cpuid is not correct
First Bad Commit
Relevant log output
✔ Container tabbyml-llamacpp-1 Created 0.0s
Attaching to llamacpp-1
llamacpp-1 | ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no
llamacpp-1 | ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
llamacpp-1 | ggml_cuda_init: found 1 CUDA devices:
llamacpp-1 | Device 0: NVIDIA GeForce GTX 1070, compute capability 6.1, VMM: yes
llamacpp-1 | warn: LLAMA_ARG_HOST environment variable is set, but will be overwritten by command line argument --host
llamacpp-1 | build: 4932 (9ffcc9e3) with cc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0 for x86_64-linux-gnu
llamacpp-1 | system info: n_threads = 4, n_threads_batch = 4, total_threads = 8
llamacpp-1 |
llamacpp-1 | system_info: n_threads = 4 (n_threads_batch = 4) / 8 | CUDA : ARCHS = 500,610,700,750,800 | USE_GRAPHS = 1 | PEER_MAX_BATCH_SIZE = 128 | CPU : SSE3 = 1 | SSSE3 = 1 | AVX = 1 | AVX2 = 1 | F16C = 1 | FMA = 1 | BMI2 = 1 | LLAMAFILE = 1 | OPENMP = 1 | AARCH64_REPACK = 1 |
llamacpp-1 |
llamacpp-1 | main: HTTP server is listening, hostname: 0.0.0.0, port: 80, http threads: 7
llamacpp-1 | main: loading model
llamacpp-1 | srv load_model: loading model '/models/Llama-3.2-3B-Instruct-Q6_K_L.gguf'
llamacpp-1 | llama_model_load_from_file_impl: using device CUDA0 (NVIDIA GeForce GTX 1070) - 8027 MiB free
llamacpp-1 | llama_model_loader: loaded meta data with 35 key-value pairs and 255 tensors from /models/Llama-3.2-3B-Instruct-Q6_K_L.gguf (version GGUF V3 (latest))
llamacpp-1 | llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
llamacpp-1 | llama_model_loader: - kv 0: general.architecture str = llama
llamacpp-1 | llama_model_loader: - kv 1: general.type str = model
llamacpp-1 | llama_model_loader: - kv 2: general.name str = Llama 3.2 3B Instruct
llamacpp-1 | llama_model_loader: - kv 3: general.finetune str = Instruct
llamacpp-1 | llama_model_loader: - kv 4: general.basename str = Llama-3.2
llamacpp-1 | llama_model_loader: - kv 5: general.size_label str = 3B
llamacpp-1 | llama_model_loader: - kv 6: general.license str = llama3.2
llamacpp-1 | llama_model_loader: - kv 7: general.tags arr[str,6] = ["facebook", "meta", "pytorch", "llam...
llamacpp-1 | llama_model_loader: - kv 8: general.languages arr[str,8] = ["en", "de", "fr", "it", "pt", "hi", ...
llamacpp-1 | llama_model_loader: - kv 9: llama.block_count u32 = 28
llamacpp-1 | llama_model_loader: - kv 10: llama.context_length u32 = 131072
llamacpp-1 | llama_model_loader: - kv 11: llama.embedding_length u32 = 3072
llamacpp-1 | llama_model_loader: - kv 12: llama.feed_forward_length u32 = 8192
llamacpp-1 | llama_model_loader: - kv 13: llama.attention.head_count u32 = 24
llamacpp-1 | llama_model_loader: - kv 14: llama.attention.head_count_kv u32 = 8
llamacpp-1 | llama_model_loader: - kv 15: llama.rope.freq_base f32 = 500000.000000
llamacpp-1 | llama_model_loader: - kv 16: llama.attention.layer_norm_rms_epsilon f32 = 0.000010
llamacpp-1 | llama_model_loader: - kv 17: llama.attention.key_length u32 = 128
llamacpp-1 | llama_model_loader: - kv 18: llama.attention.value_length u32 = 128
llamacpp-1 | llama_model_loader: - kv 19: general.file_type u32 = 18
llamacpp-1 | llama_model_loader: - kv 20: llama.vocab_size u32 = 128256
llamacpp-1 | llama_model_loader: - kv 21: llama.rope.dimension_count u32 = 128
llamacpp-1 | llama_model_loader: - kv 22: tokenizer.ggml.model str = gpt2
llamacpp-1 | llama_model_loader: - kv 23: tokenizer.ggml.pre str = llama-bpe
llamacpp-1 | llama_model_loader: - kv 24: tokenizer.ggml.tokens arr[str,128256] = ["!", "\"", "#", "$", "%", "&", "'", ...
llamacpp-1 | llama_model_loader: - kv 25: tokenizer.ggml.token_type arr[i32,128256] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
llamacpp-1 | llama_model_loader: - kv 26: tokenizer.ggml.merges arr[str,280147] = ["Ġ Ġ", "Ġ ĠĠĠ", "ĠĠ ĠĠ", "...
llamacpp-1 | llama_model_loader: - kv 27: tokenizer.ggml.bos_token_id u32 = 128000
llamacpp-1 | llama_model_loader: - kv 28: tokenizer.ggml.eos_token_id u32 = 128009
llamacpp-1 | llama_model_loader: - kv 29: tokenizer.chat_template str = {{- bos_token }}\n{%- if custom_tools ...
llamacpp-1 | llama_model_loader: - kv 30: general.quantization_version u32 = 2
llamacpp-1 | llama_model_loader: - kv 31: quantize.imatrix.file str = /models_out/Llama-3.2-3B-Instruct-GGU...
llamacpp-1 | llama_model_loader: - kv 32: quantize.imatrix.dataset str = /training_dir/calibration_datav3.txt
llamacpp-1 | llama_model_loader: - kv 33: quantize.imatrix.entries_count i32 = 196
llamacpp-1 | llama_model_loader: - kv 34: quantize.imatrix.chunks_count i32 = 125
llamacpp-1 | llama_model_loader: - type f32: 58 tensors
llamacpp-1 | llama_model_loader: - type q8_0: 1 tensors
llamacpp-1 | llama_model_loader: - type q6_K: 196 tensors
llamacpp-1 | print_info: file format = GGUF V3 (latest)
llamacpp-1 | print_info: file type = Q6_K
llamacpp-1 | print_info: file size = 2.54 GiB (6.80 BPW)
llamacpp-1 | init_tokenizer: initializing tokenizer for type 2
llamacpp-1 | load: control token: 128254 '<|reserved_special_token_246|>' is not marked as EOG
llamacpp-1 | load: control token: 128249 '<|reserved_special_token_241|>' is not marked as EOG
llamacpp-1 | load: control token: 128246 '<|reserved_special_token_238|>' is not marked as EOG
llamacpp-1 | load: control token: 128243 '<|reserved_special_token_235|>' is not marked as EOG
llamacpp-1 | load: control token: 128242 '<|reserved_special_token_234|>' is not marked as EOG
llamacpp-1 | load: control token: 128241 '<|reserved_special_token_233|>' is not marked as EOG
llamacpp-1 | load: control token: 128240 '<|reserved_special_token_232|>' is not marked as EOG
llamacpp-1 | load: control token: 128235 '<|reserved_special_token_227|>' is not marked as EOG
llamacpp-1 | load: control token: 128231 '<|reserved_special_token_223|>' is not marked as EOG
llamacpp-1 | load: control token: 128230 '<|reserved_special_token_222|>' is not marked as EOG
llamacpp-1 | load: control token: 128228 '<|reserved_special_token_220|>' is not marked as EOG
llamacpp-1 | load: control token: 128225 '<|reserved_special_token_217|>' is not marked as EOG
llamacpp-1 | load: control token: 128218 '<|reserved_special_token_210|>' is not marked as EOG
llamacpp-1 | load: control token: 128214 '<|reserved_special_token_206|>' is not marked as EOG
llamacpp-1 | load: control token: 128213 '<|reserved_special_token_205|>' is not marked as EOG
llamacpp-1 | load: control token: 128207 '<|reserved_special_token_199|>' is not marked as EOG
llamacpp-1 | load: control token: 128206 '<|reserved_special_token_198|>' is not marked as EOG
llamacpp-1 | load: control token: 128204 '<|reserved_special_token_196|>' is not marked as EOG
llamacpp-1 | load: control token: 128200 '<|reserved_special_token_192|>' is not marked as EOG
llamacpp-1 | load: control token: 128199 '<|reserved_special_token_191|>' is not marked as EOG
llamacpp-1 | load: control token: 128198 '<|reserved_special_token_190|>' is not marked as EOG
llamacpp-1 | load: control token: 128196 '<|reserved_special_token_188|>' is not marked as EOG
llamacpp-1 | load: control token: 128194 '<|reserved_special_token_186|>' is not marked as EOG
llamacpp-1 | load: control token: 128193 '<|reserved_special_token_185|>' is not marked as EOG
llamacpp-1 | load: control token: 128188 '<|reserved_special_token_180|>' is not marked as EOG
llamacpp-1 | load: control token: 128187 '<|reserved_special_token_179|>' is not marked as EOG
llamacpp-1 | load: control token: 128185 '<|reserved_special_token_177|>' is not marked as EOG
llamacpp-1 | load: control token: 128184 '<|reserved_special_token_176|>' is not marked as EOG
llamacpp-1 | load: control token: 128180 '<|reserved_special_token_172|>' is not marked as EOG
llamacpp-1 | load: control token: 128179 '<|reserved_special_token_171|>' is not marked as EOG
llamacpp-1 | load: control token: 128178 '<|reserved_special_token_170|>' is not marked as EOG
llamacpp-1 | load: control token: 128177 '<|reserved_special_token_169|>' is not marked as EOG
llamacpp-1 | load: control token: 128176 '<|reserved_special_token_168|>' is not marked as EOG
llamacpp-1 | load: control token: 128175 '<|reserved_special_token_167|>' is not marked as EOG
llamacpp-1 | load: control token: 128171 '<|reserved_special_token_163|>' is not marked as EOG
llamacpp-1 | load: control token: 128170 '<|reserved_special_token_162|>' is not marked as EOG
llamacpp-1 | load: control token: 128169 '<|reserved_special_token_161|>' is not marked as EOG
llamacpp-1 | load: control token: 128168 '<|reserved_special_token_160|>' is not marked as EOG
llamacpp-1 | load: control token: 128165 '<|reserved_special_token_157|>' is not marked as EOG
llamacpp-1 | load: control token: 128162 '<|reserved_special_token_154|>' is not marked as EOG
llamacpp-1 | load: control token: 128158 '<|reserved_special_token_150|>' is not marked as EOG
llamacpp-1 | load: control token: 128156 '<|reserved_special_token_148|>' is not marked as EOG
llamacpp-1 | load: control token: 128155 '<|reserved_special_token_147|>' is not marked as EOG
llamacpp-1 | load: control token: 128154 '<|reserved_special_token_146|>' is not marked as EOG
llamacpp-1 | load: control token: 128151 '<|reserved_special_token_143|>' is not marked as EOG
llamacpp-1 | load: control token: 128149 '<|reserved_special_token_141|>' is not marked as EOG
llamacpp-1 | load: control token: 128147 '<|reserved_special_token_139|>' is not marked as EOG
llamacpp-1 | load: control token: 128146 '<|reserved_special_token_138|>' is not marked as EOG
llamacpp-1 | load: control token: 128144 '<|reserved_special_token_136|>' is not marked as EOG
llamacpp-1 | load: control token: 128142 '<|reserved_special_token_134|>' is not marked as EOG
llamacpp-1 | load: control token: 128141 '<|reserved_special_token_133|>' is not marked as EOG
llamacpp-1 | load: control token: 128138 '<|reserved_special_token_130|>' is not marked as EOG
llamacpp-1 | load: control token: 128136 '<|reserved_special_token_128|>' is not marked as EOG
llamacpp-1 | load: control token: 128135 '<|reserved_special_token_127|>' is not marked as EOG
llamacpp-1 | load: control token: 128134 '<|reserved_special_token_126|>' is not marked as EOG
llamacpp-1 | load: control token: 128133 '<|reserved_special_token_125|>' is not marked as EOG
llamacpp-1 | load: control token: 128131 '<|reserved_special_token_123|>' is not marked as EOG
llamacpp-1 | load: control token: 128128 '<|reserved_special_token_120|>' is not marked as EOG
llamacpp-1 | load: control token: 128124 '<|reserved_special_token_116|>' is not marked as EOG
llamacpp-1 | load: control token: 128123 '<|reserved_special_token_115|>' is not marked as EOG
llamacpp-1 | load: control token: 128122 '<|reserved_special_token_114|>' is not marked as EOG
llamacpp-1 | load: control token: 128119 '<|reserved_special_token_111|>' is not marked as EOG
llamacpp-1 | load: control token: 128115 '<|reserved_special_token_107|>' is not marked as EOG
llamacpp-1 | load: control token: 128112 '<|reserved_special_token_104|>' is not marked as EOG
llamacpp-1 | load: control token: 128110 '<|reserved_special_token_102|>' is not marked as EOG
llamacpp-1 | load: control token: 128109 '<|reserved_special_token_101|>' is not marked as EOG
llamacpp-1 | load: control token: 128108 '<|reserved_special_token_100|>' is not marked as EOG
llamacpp-1 | load: control token: 128106 '<|reserved_special_token_98|>' is not marked as EOG
llamacpp-1 | load: control token: 128103 '<|reserved_special_token_95|>' is not marked as EOG
llamacpp-1 | load: control token: 128102 '<|reserved_special_token_94|>' is not marked as EOG
llamacpp-1 | load: control token: 128101 '<|reserved_special_token_93|>' is not marked as EOG
llamacpp-1 | load: control token: 128097 '<|reserved_special_token_89|>' is not marked as EOG
llamacpp-1 | load: control token: 128091 '<|reserved_special_token_83|>' is not marked as EOG
llamacpp-1 | load: control token: 128090 '<|reserved_special_token_82|>' is not marked as EOG
llamacpp-1 | load: control token: 128089 '<|reserved_special_token_81|>' is not marked as EOG
llamacpp-1 | load: control token: 128087 '<|reserved_special_token_79|>' is not marked as EOG
llamacpp-1 | load: control token: 128085 '<|reserved_special_token_77|>' is not marked as EOG
llamacpp-1 | load: control token: 128081 '<|reserved_special_token_73|>' is not marked as EOG
llamacpp-1 | load: control token: 128078 '<|reserved_special_token_70|>' is not marked as EOG
llamacpp-1 | load: control token: 128076 '<|reserved_special_token_68|>' is not marked as EOG
llamacpp-1 | load: control token: 128075 '<|reserved_special_token_67|>' is not marked as EOG
llamacpp-1 | load: control token: 128073 '<|reserved_special_token_65|>' is not marked as EOG
llamacpp-1 | load: control token: 128068 '<|reserved_special_token_60|>' is not marked as EOG
llamacpp-1 | load: control token: 128067 '<|reserved_special_token_59|>' is not marked as EOG
llamacpp-1 | load: control token: 128065 '<|reserved_special_token_57|>' is not marked as EOG
llamacpp-1 | load: control token: 128063 '<|reserved_special_token_55|>' is not marked as EOG
llamacpp-1 | load: control token: 128062 '<|reserved_special_token_54|>' is not marked as EOG
llamacpp-1 | load: control token: 128060 '<|reserved_special_token_52|>' is not marked as EOG
llamacpp-1 | load: control token: 128059 '<|reserved_special_token_51|>' is not marked as EOG
llamacpp-1 | load: control token: 128057 '<|reserved_special_token_49|>' is not marked as EOG
llamacpp-1 | load: control token: 128054 '<|reserved_special_token_46|>' is not marked as EOG
llamacpp-1 | load: control token: 128046 '<|reserved_special_token_38|>' is not marked as EOG
llamacpp-1 | load: control token: 128045 '<|reserved_special_token_37|>' is not marked as EOG
llamacpp-1 | load: control token: 128044 '<|reserved_special_token_36|>' is not marked as EOG
llamacpp-1 | load: control token: 128043 '<|reserved_special_token_35|>' is not marked as EOG
llamacpp-1 | load: control token: 128038 '<|reserved_special_token_30|>' is not marked as EOG
llamacpp-1 | load: control token: 128036 '<|reserved_special_token_28|>' is not marked as EOG
llamacpp-1 | load: control token: 128035 '<|reserved_special_token_27|>' is not marked as EOG
llamacpp-1 | load: control token: 128032 '<|reserved_special_token_24|>' is not marked as EOG
llamacpp-1 | load: control token: 128028 '<|reserved_special_token_20|>' is not marked as EOG
llamacpp-1 | load: control token: 128027 '<|reserved_special_token_19|>' is not marked as EOG
llamacpp-1 | load: control token: 128024 '<|reserved_special_token_16|>' is not marked as EOG
llamacpp-1 | load: control token: 128023 '<|reserved_special_token_15|>' is not marked as EOG
llamacpp-1 | load: control token: 128022 '<|reserved_special_token_14|>' is not marked as EOG
llamacpp-1 | load: control token: 128021 '<|reserved_special_token_13|>' is not marked as EOG
llamacpp-1 | load: control token: 128018 '<|reserved_special_token_10|>' is not marked as EOG
llamacpp-1 | load: control token: 128016 '<|reserved_special_token_8|>' is not marked as EOG
llamacpp-1 | load: control token: 128015 '<|reserved_special_token_7|>' is not marked as EOG
llamacpp-1 | load: control token: 128013 '<|reserved_special_token_5|>' is not marked as EOG
llamacpp-1 | load: control token: 128011 '<|reserved_special_token_3|>' is not marked as EOG
llamacpp-1 | load: control token: 128005 '<|reserved_special_token_2|>' is not marked as EOG
llamacpp-1 | load: control token: 128004 '<|finetune_right_pad_id|>' is not marked as EOG
llamacpp-1 | load: control token: 128002 '<|reserved_special_token_0|>' is not marked as EOG
llamacpp-1 | load: control token: 128252 '<|reserved_special_token_244|>' is not marked as EOG
llamacpp-1 | load: control token: 128190 '<|reserved_special_token_182|>' is not marked as EOG
llamacpp-1 | load: control token: 128183 '<|reserved_special_token_175|>' is not marked as EOG
llamacpp-1 | load: control token: 128137 '<|reserved_special_token_129|>' is not marked as EOG
llamacpp-1 | load: control token: 128182 '<|reserved_special_token_174|>' is not marked as EOG
llamacpp-1 | load: control token: 128040 '<|reserved_special_token_32|>' is not marked as EOG
llamacpp-1 | load: control token: 128048 '<|reserved_special_token_40|>' is not marked as EOG
llamacpp-1 | load: control token: 128092 '<|reserved_special_token_84|>' is not marked as EOG
llamacpp-1 | load: control token: 128215 '<|reserved_special_token_207|>' is not marked as EOG
llamacpp-1 | load: control token: 128107 '<|reserved_special_token_99|>' is not marked as EOG
llamacpp-1 | load: control token: 128208 '<|reserved_special_token_200|>' is not marked as EOG
llamacpp-1 | load: control token: 128145 '<|reserved_special_token_137|>' is not marked as EOG
llamacpp-1 | load: control token: 128031 '<|reserved_special_token_23|>' is not marked as EOG
llamacpp-1 | load: control token: 128129 '<|reserved_special_token_121|>' is not marked as EOG
llamacpp-1 | load: control token: 128201 '<|reserved_special_token_193|>' is not marked as EOG
llamacpp-1 | load: control token: 128074 '<|reserved_special_token_66|>' is not marked as EOG
llamacpp-1 | load: control token: 128095 '<|reserved_special_token_87|>' is not marked as EOG
llamacpp-1 | load: control token: 128186 '<|reserved_special_token_178|>' is not marked as EOG
llamacpp-1 | load: control token: 128143 '<|reserved_special_token_135|>' is not marked as EOG
llamacpp-1 | load: control token: 128229 '<|reserved_special_token_221|>' is not marked as EOG
llamacpp-1 | load: control token: 128007 '<|end_header_id|>' is not marked as EOG
llamacpp-1 | load: control token: 128055 '<|reserved_special_token_47|>' is not marked as EOG
llamacpp-1 | load: control token: 128056 '<|reserved_special_token_48|>' is not marked as EOG
llamacpp-1 | load: control token: 128061 '<|reserved_special_token_53|>' is not marked as EOG
llamacpp-1 | load: control token: 128153 '<|reserved_special_token_145|>' is not marked as EOG
llamacpp-1 | load: control token: 128152 '<|reserved_special_token_144|>' is not marked as EOG
llamacpp-1 | load: control token: 128212 '<|reserved_special_token_204|>' is not marked as EOG
llamacpp-1 | load: control token: 128172 '<|reserved_special_token_164|>' is not marked as EOG
llamacpp-1 | load: control token: 128160 '<|reserved_special_token_152|>' is not marked as EOG
llamacpp-1 | load: control token: 128041 '<|reserved_special_token_33|>' is not marked as EOG
llamacpp-1 | load: control token: 128181 '<|reserved_special_token_173|>' is not marked as EOG
llamacpp-1 | load: control token: 128094 '<|reserved_special_token_86|>' is not marked as EOG
llamacpp-1 | load: control token: 128118 '<|reserved_special_token_110|>' is not marked as EOG
llamacpp-1 | load: control token: 128236 '<|reserved_special_token_228|>' is not marked as EOG
llamacpp-1 | load: control token: 128148 '<|reserved_special_token_140|>' is not marked as EOG
llamacpp-1 | load: control token: 128042 '<|reserved_special_token_34|>' is not marked as EOG
llamacpp-1 | load: control token: 128139 '<|reserved_special_token_131|>' is not marked as EOG
llamacpp-1 | load: control token: 128173 '<|reserved_special_token_165|>' is not marked as EOG
llamacpp-1 | load: control token: 128239 '<|reserved_special_token_231|>' is not marked as EOG
llamacpp-1 | load: control token: 128157 '<|reserved_special_token_149|>' is not marked as EOG
llamacpp-1 | load: control token: 128052 '<|reserved_special_token_44|>' is not marked as EOG
llamacpp-1 | load: control token: 128026 '<|reserved_special_token_18|>' is not marked as EOG
llamacpp-1 | load: control token: 128003 '<|reserved_special_token_1|>' is not marked as EOG
llamacpp-1 | load: control token: 128019 '<|reserved_special_token_11|>' is not marked as EOG
llamacpp-1 | load: control token: 128116 '<|reserved_special_token_108|>' is not marked as EOG
llamacpp-1 | load: control token: 128161 '<|reserved_special_token_153|>' is not marked as EOG
llamacpp-1 | load: control token: 128226 '<|reserved_special_token_218|>' is not marked as EOG
llamacpp-1 | load: control token: 128159 '<|reserved_special_token_151|>' is not marked as EOG
llamacpp-1 | load: control token: 128012 '<|reserved_special_token_4|>' is not marked as EOG
llamacpp-1 | load: control token: 128088 '<|reserved_special_token_80|>' is not marked as EOG
llamacpp-1 | load: control token: 128163 '<|reserved_special_token_155|>' is not marked as EOG
llamacpp-1 | load: control token: 128001 '<|end_of_text|>' is not marked as EOG
llamacpp-1 | load: control token: 128113 '<|reserved_special_token_105|>' is not marked as EOG
llamacpp-1 | load: control token: 128250 '<|reserved_special_token_242|>' is not marked as EOG
llamacpp-1 | load: control token: 128125 '<|reserved_special_token_117|>' is not marked as EOG
llamacpp-1 | load: control token: 128053 '<|reserved_special_token_45|>' is not marked as EOG
llamacpp-1 | load: control token: 128224 '<|reserved_special_token_216|>' is not marked as EOG
llamacpp-1 | load: control token: 128247 '<|reserved_special_token_239|>' is not marked as EOG
llamacpp-1 | load: control token: 128251 '<|reserved_special_token_243|>' is not marked as EOG
llamacpp-1 | load: control token: 128216 '<|reserved_special_token_208|>' is not marked as EOG
llamacpp-1 | load: control token: 128006 '<|start_header_id|>' is not marked as EOG
llamacpp-1 | load: control token: 128211 '<|reserved_special_token_203|>' is not marked as EOG
llamacpp-1 | load: control token: 128077 '<|reserved_special_token_69|>' is not marked as EOG
llamacpp-1 | load: control token: 128237 '<|reserved_special_token_229|>' is not marked as EOG
llamacpp-1 | load: control token: 128086 '<|reserved_special_token_78|>' is not marked as EOG
llamacpp-1 | load: control token: 128227 '<|reserved_special_token_219|>' is not marked as EOG
llamacpp-1 | load: control token: 128058 '<|reserved_special_token_50|>' is not marked as EOG
llamacpp-1 | load: control token: 128100 '<|reserved_special_token_92|>' is not marked as EOG
llamacpp-1 | load: control token: 128209 '<|reserved_special_token_201|>' is not marked as EOG
llamacpp-1 | load: control token: 128084 '<|reserved_special_token_76|>' is not marked as EOG
llamacpp-1 | load: control token: 128071 '<|reserved_special_token_63|>' is not marked as EOG
llamacpp-1 | load: control token: 128070 '<|reserved_special_token_62|>' is not marked as EOG
llamacpp-1 | load: control token: 128049 '<|reserved_special_token_41|>' is not marked as EOG
llamacpp-1 | load: control token: 128197 '<|reserved_special_token_189|>' is not marked as EOG
llamacpp-1 | load: control token: 128072 '<|reserved_special_token_64|>' is not marked as EOG
llamacpp-1 | load: control token: 128000 '<|begin_of_text|>' is not marked as EOG
llamacpp-1 | load: control token: 128223 '<|reserved_special_token_215|>' is not marked as EOG
llamacpp-1 | load: control token: 128217 '<|reserved_special_token_209|>' is not marked as EOG
llamacpp-1 | load: control token: 128111 '<|reserved_special_token_103|>' is not marked as EOG
llamacpp-1 | load: control token: 128203 '<|reserved_special_token_195|>' is not marked as EOG
llamacpp-1 | load: control token: 128051 '<|reserved_special_token_43|>' is not marked as EOG
llamacpp-1 | load: control token: 128030 '<|reserved_special_token_22|>' is not marked as EOG
llamacpp-1 | load: control token: 128117 '<|reserved_special_token_109|>' is not marked as EOG
llamacpp-1 | load: control token: 128010 '<|python_tag|>' is not marked as EOG
llamacpp-1 | load: control token: 128238 '<|reserved_special_token_230|>' is not marked as EOG
llamacpp-1 | load: control token: 128255 '<|reserved_special_token_247|>' is not marked as EOG
llamacpp-1 | load: control token: 128202 '<|reserved_special_token_194|>' is not marked as EOG
llamacpp-1 | load: control token: 128132 '<|reserved_special_token_124|>' is not marked as EOG
llamacpp-1 | load: control token: 128248 '<|reserved_special_token_240|>' is not marked as EOG
llamacpp-1 | load: control token: 128167 '<|reserved_special_token_159|>' is not marked as EOG
llamacpp-1 | load: control token: 128127 '<|reserved_special_token_119|>' is not marked as EOG
llamacpp-1 | load: control token: 128105 '<|reserved_special_token_97|>' is not marked as EOG
llamacpp-1 | load: control token: 128039 '<|reserved_special_token_31|>' is not marked as EOG
llamacpp-1 | load: control token: 128232 '<|reserved_special_token_224|>' is not marked as EOG
llamacpp-1 | load: control token: 128166 '<|reserved_special_token_158|>' is not marked as EOG
llamacpp-1 | load: control token: 128130 '<|reserved_special_token_122|>' is not marked as EOG
llamacpp-1 | load: control token: 128114 '<|reserved_special_token_106|>' is not marked as EOG
llamacpp-1 | load: control token: 128234 '<|reserved_special_token_226|>' is not marked as EOG
llamacpp-1 | load: control token: 128191 '<|reserved_special_token_183|>' is not marked as EOG
llamacpp-1 | load: control token: 128064 '<|reserved_special_token_56|>' is not marked as EOG
llamacpp-1 | load: control token: 128140 '<|reserved_special_token_132|>' is not marked as EOG
llamacpp-1 | load: control token: 128096 '<|reserved_special_token_88|>' is not marked as EOG
llamacpp-1 | load: control token: 128098 '<|reserved_special_token_90|>' is not marked as EOG
llamacpp-1 | load: control token: 128192 '<|reserved_special_token_184|>' is not marked as EOG
llamacpp-1 | load: control token: 128093 '<|reserved_special_token_85|>' is not marked as EOG
llamacpp-1 | load: control token: 128150 '<|reserved_special_token_142|>' is not marked as EOG
llamacpp-1 | load: control token: 128222 '<|reserved_special_token_214|>' is not marked as EOG
llamacpp-1 | load: control token: 128233 '<|reserved_special_token_225|>' is not marked as EOG
llamacpp-1 | load: control token: 128220 '<|reserved_special_token_212|>' is not marked as EOG
llamacpp-1 | load: control token: 128034 '<|reserved_special_token_26|>' is not marked as EOG
llamacpp-1 | load: control token: 128033 '<|reserved_special_token_25|>' is not marked as EOG
llamacpp-1 | load: control token: 128253 '<|reserved_special_token_245|>' is not marked as EOG
llamacpp-1 | load: control token: 128195 '<|reserved_special_token_187|>' is not marked as EOG
llamacpp-1 | load: control token: 128099 '<|reserved_special_token_91|>' is not marked as EOG
llamacpp-1 | load: control token: 128189 '<|reserved_special_token_181|>' is not marked as EOG
llamacpp-1 | load: control token: 128210 '<|reserved_special_token_202|>' is not marked as EOG
llamacpp-1 | load: control token: 128174 '<|reserved_special_token_166|>' is not marked as EOG
llamacpp-1 | load: control token: 128083 '<|reserved_special_token_75|>' is not marked as EOG
llamacpp-1 | load: control token: 128080 '<|reserved_special_token_72|>' is not marked as EOG
llamacpp-1 | load: control token: 128104 '<|reserved_special_token_96|>' is not marked as EOG
llamacpp-1 | load: control token: 128082 '<|reserved_special_token_74|>' is not marked as EOG
llamacpp-1 | load: control token: 128219 '<|reserved_special_token_211|>' is not marked as EOG
llamacpp-1 | load: control token: 128017 '<|reserved_special_token_9|>' is not marked as EOG
llamacpp-1 | load: control token: 128050 '<|reserved_special_token_42|>' is not marked as EOG
llamacpp-1 | load: control token: 128205 '<|reserved_special_token_197|>' is not marked as EOG
llamacpp-1 | load: control token: 128047 '<|reserved_special_token_39|>' is not marked as EOG
llamacpp-1 | load: control token: 128164 '<|reserved_special_token_156|>' is not marked as EOG
llamacpp-1 | load: control token: 128020 '<|reserved_special_token_12|>' is not marked as EOG
llamacpp-1 | load: control token: 128069 '<|reserved_special_token_61|>' is not marked as EOG
llamacpp-1 | load: control token: 128245 '<|reserved_special_token_237|>' is not marked as EOG
llamacpp-1 | load: control token: 128121 '<|reserved_special_token_113|>' is not marked as EOG
llamacpp-1 | load: control token: 128079 '<|reserved_special_token_71|>' is not marked as EOG
llamacpp-1 | load: control token: 128037 '<|reserved_special_token_29|>' is not marked as EOG
llamacpp-1 | load: control token: 128244 '<|reserved_special_token_236|>' is not marked as EOG
llamacpp-1 | load: control token: 128029 '<|reserved_special_token_21|>' is not marked as EOG
llamacpp-1 | load: control token: 128221 '<|reserved_special_token_213|>' is not marked as EOG
llamacpp-1 | load: control token: 128066 '<|reserved_special_token_58|>' is not marked as EOG
llamacpp-1 | load: control token: 128120 '<|reserved_special_token_112|>' is not marked as EOG
llamacpp-1 | load: control token: 128014 '<|reserved_special_token_6|>' is not marked as EOG
llamacpp-1 | load: control token: 128025 '<|reserved_special_token_17|>' is not marked as EOG
llamacpp-1 | load: control token: 128126 '<|reserved_special_token_118|>' is not marked as EOG
llamacpp-1 | load: special tokens cache size = 256
llamacpp-1 | load: token to piece cache size = 0.7999 MB
llamacpp-1 | print_info: arch = llama
llamacpp-1 | print_info: vocab_only = 0
llamacpp-1 | print_info: n_ctx_train = 131072
llamacpp-1 | print_info: n_embd = 3072
llamacpp-1 | print_info: n_layer = 28
llamacpp-1 | print_info: n_head = 24
llamacpp-1 | print_info: n_head_kv = 8
llamacpp-1 | print_info: n_rot = 128
llamacpp-1 | print_info: n_swa = 0
llamacpp-1 | print_info: n_swa_pattern = 1
llamacpp-1 | print_info: n_embd_head_k = 128
llamacpp-1 | print_info: n_embd_head_v = 128
llamacpp-1 | print_info: n_gqa = 3
llamacpp-1 | print_info: n_embd_k_gqa = 1024
llamacpp-1 | print_info: n_embd_v_gqa = 1024
llamacpp-1 | print_info: f_norm_eps = 0.0e+00
llamacpp-1 | print_info: f_norm_rms_eps = 1.0e-05
llamacpp-1 | print_info: f_clamp_kqv = 0.0e+00
llamacpp-1 | print_info: f_max_alibi_bias = 0.0e+00
llamacpp-1 | print_info: f_logit_scale = 0.0e+00
llamacpp-1 | print_info: f_attn_scale = 0.0e+00
llamacpp-1 | print_info: n_ff = 8192
llamacpp-1 | print_info: n_expert = 0
llamacpp-1 | print_info: n_expert_used = 0
llamacpp-1 | print_info: causal attn = 1
llamacpp-1 | print_info: pooling type = 0
llamacpp-1 | print_info: rope type = 0
llamacpp-1 | print_info: rope scaling = linear
llamacpp-1 | print_info: freq_base_train = 500000.0
llamacpp-1 | print_info: freq_scale_train = 1
llamacpp-1 | print_info: n_ctx_orig_yarn = 131072
llamacpp-1 | print_info: rope_finetuned = unknown
llamacpp-1 | print_info: ssm_d_conv = 0
llamacpp-1 | print_info: ssm_d_inner = 0
llamacpp-1 | print_info: ssm_d_state = 0
llamacpp-1 | print_info: ssm_dt_rank = 0
llamacpp-1 | print_info: ssm_dt_b_c_rms = 0
llamacpp-1 | print_info: model type = 3B
llamacpp-1 | print_info: model params = 3.21 B
llamacpp-1 | print_info: general.name = Llama 3.2 3B Instruct
llamacpp-1 | print_info: vocab type = BPE
llamacpp-1 | print_info: n_vocab = 128256
llamacpp-1 | print_info: n_merges = 280147
llamacpp-1 | print_info: BOS token = 128000 '<|begin_of_text|>'
llamacpp-1 | print_info: EOS token = 128009 '<|eot_id|>'
llamacpp-1 | print_info: EOT token = 128009 '<|eot_id|>'
llamacpp-1 | print_info: EOM token = 128008 '<|eom_id|>'
llamacpp-1 | print_info: LF token = 198 'Ċ'
llamacpp-1 | print_info: EOG token = 128008 '<|eom_id|>'
llamacpp-1 | print_info: EOG token = 128009 '<|eot_id|>'
llamacpp-1 | print_info: max token length = 256
llamacpp-1 | load_tensors: loading model tensors, this can take a while... (mmap = true)
llamacpp-1 | load_tensors: layer 0 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 1 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 2 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 3 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 4 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 5 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 6 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 7 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 8 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 9 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 10 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 11 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 12 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 13 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 14 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 15 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 16 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 17 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 18 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 19 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 20 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 21 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 22 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 23 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 24 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 25 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 26 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 27 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: layer 28 assigned to device CUDA0, is_swa = 0
llamacpp-1 | load_tensors: tensor 'token_embd.weight' (q8_0) (and 0 others) cannot be used with preferred buffer type CPU_AARCH64, using CPU instead
llamacpp-1 | load_tensors: offloading 28 repeating layers to GPU
llamacpp-1 | load_tensors: offloading output layer to GPU
llamacpp-1 | load_tensors: offloaded 29/29 layers to GPU
llamacpp-1 | load_tensors: CUDA0 model buffer size = 2604.90 MiB
llamacpp-1 | load_tensors: CPU_Mapped model buffer size = 399.23 MiB
llamacpp-1 | ...........................................................................
llamacpp-1 | llama_context: constructing llama_context
llamacpp-1 | llama_context: n_seq_max = 1
llamacpp-1 | llama_context: n_ctx = 8192
llamacpp-1 | llama_context: n_ctx_per_seq = 8192
llamacpp-1 | llama_context: n_batch = 2048
llamacpp-1 | llama_context: n_ubatch = 512
llamacpp-1 | llama_context: causal_attn = 1
llamacpp-1 | llama_context: flash_attn = 0
llamacpp-1 | llama_context: freq_base = 500000.0
llamacpp-1 | llama_context: freq_scale = 1
llamacpp-1 | llama_context: n_ctx_per_seq (8192) < n_ctx_train (131072) -- the full capacity of the model will not be utilized
llamacpp-1 | set_abort_callback: call
llamacpp-1 | llama_context: CUDA_Host output buffer size = 0.49 MiB
llamacpp-1 | llama_context: n_ctx = 8192
llamacpp-1 | llama_context: n_ctx = 8192 (padded)
llamacpp-1 | init: kv_size = 8192, offload = 1, type_k = 'f16', type_v = 'f16', n_layer = 28, can_shift = 1
llamacpp-1 | init: layer 0: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 1: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 2: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 3: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 4: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 5: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 6: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 7: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 8: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 9: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 10: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 11: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 12: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 13: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 14: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 15: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 16: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 17: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 18: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 19: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 20: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 21: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 22: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 23: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 24: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 25: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 26: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: layer 27: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024, dev = CUDA0
llamacpp-1 | init: CUDA0 KV buffer size = 896.00 MiB
llamacpp-1 | llama_context: KV self size = 896.00 MiB, K (f16): 448.00 MiB, V (f16): 448.00 MiB
llamacpp-1 | llama_context: enumerating backends
llamacpp-1 | llama_context: backend_ptrs.size() = 2
llamacpp-1 | llama_context: max_nodes = 65536
llamacpp-1 | llama_context: n_tokens = 512, n_seqs = 1, n_outputs = 512
llamacpp-1 | llama_context: CUDA0 compute buffer size = 424.00 MiB
llamacpp-1 | llama_context: CUDA_Host compute buffer size = 22.01 MiB
llamacpp-1 | llama_context: graph nodes = 958
llamacpp-1 | llama_context: graph splits = 2
llamacpp-1 | clear_adapter_lora: call
llamacpp-1 | common_init_from_params: setting dry_penalty_last_n to ctx_size = 8192
llamacpp-1 | common_init_from_params: warming up the model with an empty run - please wait ... (--no-warmup to disable)
llamacpp-1 | set_warmup: value = 1
llamacpp-1 exited with code 0