Task 77717
| Name | minerva-7bI-q6-llama-inference_176166_1739055910.676393_6071_2 |
| Workunit | 52442 |
| Created | 14 Feb 2025, 20:27:24 UTC |
| Sent | 14 Feb 2025, 22:03:38 UTC |
| Report deadline | 21 Feb 2025, 22:03:38 UTC |
| Received | 15 Feb 2025, 0:27:17 UTC |
| Server state | Over |
| Outcome | Success |
| Client state | Done |
| Exit status | 0 (0x00000000) |
| Computer ID | 94 |
| Run time | 8 min 20 sec |
| CPU time | 32 min 46 sec |
| Validate state | Valid |
| Credit | 69.78 |
| Device peak FLOPS | 49.45 GFLOPS |
| Application version | Minerva 7B Instruct Q6 inference via LLama.cpp v4.00 (mt) x86_64-pc-linux-gnu |
| Peak working set size | 5.72 GB |
| Peak swap size | 5.80 GB |
| Peak disk usage | 45.88 KB |
Stderr output
<core_client_version>7.24.1</core_client_version>
<![CDATA[
<stderr_txt>
llama_model_loader: loaded meta data with 51 key-value pairs and 291 tensors from ../../projects/boinc.llmentor.org_LLMentorGrid/minerva-7b-instruct-v1.0-q6_k.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 = llama
llama_model_loader: - kv 1: general.type str = model
llama_model_loader: - kv 2: general.name str = Minerva 7b Inst
llama_model_loader: - kv 3: general.finetune str = inst
llama_model_loader: - kv 4: general.basename str = minerva
llama_model_loader: - kv 5: general.size_label str = 7B
llama_model_loader: - kv 6: general.license str = apache-2.0
llama_model_loader: - kv 7: general.base_model.count u32 = 1
llama_model_loader: - kv 8: general.base_model.0.name str = Minerva 7B Base v1.0
llama_model_loader: - kv 9: general.base_model.0.version str = v1.0
llama_model_loader: - kv 10: general.base_model.0.organization str = Sapienzanlp
llama_model_loader: - kv 11: general.base_model.0.repo_url str = https://huggingface.co/sapienzanlp/Mi...
llama_model_loader: - kv 12: general.dataset.count u32 = 3
llama_model_loader: - kv 13: general.dataset.0.name str = Ultrafeedback_Binarized
llama_model_loader: - kv 14: general.dataset.0.organization str = HuggingFaceH4
llama_model_loader: - kv 15: general.dataset.0.repo_url str = https://huggingface.co/HuggingFaceH4/...
llama_model_loader: - kv 16: general.dataset.1.name str = ALERT
llama_model_loader: - kv 17: general.dataset.1.organization str = Babelscape
llama_model_loader: - kv 18: general.dataset.1.repo_url str = https://huggingface.co/Babelscape/ALERT
llama_model_loader: - kv 19: general.dataset.2.name str = Evol Dpo Ita
llama_model_loader: - kv 20: general.dataset.2.organization str = Efederici
llama_model_loader: - kv 21: general.dataset.2.repo_url str = https://huggingface.co/efederici/evol...
llama_model_loader: - kv 22: general.tags arr[str,3] = ["sft", "dpo", "text-generation"]
llama_model_loader: - kv 23: general.languages arr[str,2] = ["it", "en"]
llama_model_loader: - kv 24: llama.block_count u32 = 32
llama_model_loader: - kv 25: llama.context_length u32 = 4096
llama_model_loader: - kv 26: llama.embedding_length u32 = 4096
llama_model_loader: - kv 27: llama.feed_forward_length u32 = 14336
llama_model_loader: - kv 28: llama.attention.head_count u32 = 32
llama_model_loader: - kv 29: llama.attention.head_count_kv u32 = 8
llama_model_loader: - kv 30: llama.rope.freq_base f32 = 10000.000000
llama_model_loader: - kv 31: llama.attention.layer_norm_rms_epsilon f32 = 0.000010
llama_model_loader: - kv 32: llama.attention.key_length u32 = 128
llama_model_loader: - kv 33: llama.attention.value_length u32 = 128
llama_model_loader: - kv 34: general.file_type u32 = 18
llama_model_loader: - kv 35: llama.vocab_size u32 = 51264
llama_model_loader: - kv 36: llama.rope.dimension_count u32 = 128
llama_model_loader: - kv 37: tokenizer.ggml.model str = gpt2
llama_model_loader: - kv 38: tokenizer.ggml.pre str = minerva-7b
llama_model_loader: - kv 39: tokenizer.ggml.tokens arr[str,51264] = ["<s>", "</s>", "<unk>", "!", "\"", "...
llama_model_loader: - kv 40: tokenizer.ggml.token_type arr[i32,51264] = [3, 3, 3, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
llama_model_loader: - kv 41: tokenizer.ggml.merges arr[str,50955] = ["Ä  a", "i n", "Ä  Ä ", "e r", "o n"...
llama_model_loader: - kv 42: tokenizer.ggml.bos_token_id u32 = 0
llama_model_loader: - kv 43: tokenizer.ggml.eos_token_id u32 = 51202
llama_model_loader: - kv 44: tokenizer.ggml.unknown_token_id u32 = 2
llama_model_loader: - kv 45: tokenizer.ggml.padding_token_id u32 = 51202
llama_model_loader: - kv 46: tokenizer.ggml.add_bos_token bool = true
llama_model_loader: - kv 47: tokenizer.ggml.add_eos_token bool = false
llama_model_loader: - kv 48: tokenizer.chat_template str = {% set loop_messages = messages %}{% ...
llama_model_loader: - kv 49: tokenizer.ggml.add_space_prefix bool = false
llama_model_loader: - kv 50: general.quantization_version u32 = 2
llama_model_loader: - type f32: 65 tensors
llama_model_loader: - type q6_K: 226 tensors
llm_load_vocab: control token: 51201 '<|end_header_id|>' is not marked as EOG
llm_load_vocab: control token: 51200 '<|start_header_id|>' is not marked as EOG
llm_load_vocab: control token: 0 '<s>' is not marked as EOG
llm_load_vocab: control token: 1 '</s>' is not marked as EOG
llm_load_vocab: control token: 2 '<unk>' is not marked as EOG
llm_load_vocab: special tokens cache size = 6
llm_load_vocab: token to piece cache size = 0.3264 MB
llm_load_print_meta: format = GGUF V3 (latest)
llm_load_print_meta: arch = llama
llm_load_print_meta: vocab type = BPE
llm_load_print_meta: n_vocab = 51264
llm_load_print_meta: n_merges = 50955
llm_load_print_meta: vocab_only = 0
llm_load_print_meta: n_ctx_train = 4096
llm_load_print_meta: n_embd = 4096
llm_load_print_meta: n_layer = 32
llm_load_print_meta: n_head = 32
llm_load_print_meta: n_head_kv = 8
llm_load_print_meta: n_rot = 128
llm_load_print_meta: n_swa = 0
llm_load_print_meta: n_embd_head_k = 128
llm_load_print_meta: n_embd_head_v = 128
llm_load_print_meta: n_gqa = 4
llm_load_print_meta: n_embd_k_gqa = 1024
llm_load_print_meta: n_embd_v_gqa = 1024
llm_load_print_meta: f_norm_eps = 0.0e+00
llm_load_print_meta: f_norm_rms_eps = 1.0e-05
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 = 14336
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 = 0
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 = 4096
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 = 8B
llm_load_print_meta: model ftype = Q6_K
llm_load_print_meta: model params = 7.40 B
llm_load_print_meta: model size = 5.65 GiB (6.56 BPW)
llm_load_print_meta: general.name = Minerva 7b Inst
llm_load_print_meta: BOS token = 0 '<s>'
llm_load_print_meta: EOS token = 51202 '<|eot_id|>'
llm_load_print_meta: EOT token = 51202 '<|eot_id|>'
llm_load_print_meta: UNK token = 2 '<unk>'
llm_load_print_meta: PAD token = 51202 '<|eot_id|>'
llm_load_print_meta: LF token = 129 'Ä'
llm_load_print_meta: EOG token = 51202 '<|eot_id|>'
llm_load_print_meta: max token length = 128
llm_load_tensors: tensor 'token_embd.weight' (q6_K) (and 290 others) cannot be used with preferred buffer type CPU_AARCH64, using CPU instead
llm_load_tensors: CPU_Mapped model buffer size = 5789.55 MiB
.................................................................................................
llama_new_context_with_model: n_seq_max = 1
llama_new_context_with_model: n_ctx = 64
llama_new_context_with_model: n_ctx_per_seq = 64
llama_new_context_with_model: n_batch = 32
llama_new_context_with_model: n_ubatch = 32
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_new_context_with_model: n_ctx_per_seq (64) < n_ctx_train (4096) -- the full capacity of the model will not be utilized
llama_kv_cache_init: kv_size = 64, offload = 1, type_k = 'f16', type_v = 'f16', n_layer = 32
llama_kv_cache_init: layer 0: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 1: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 2: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 3: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 4: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 5: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 6: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 7: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 8: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 9: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 10: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 11: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 12: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 13: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 14: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 15: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 16: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 17: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 18: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 19: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 20: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 21: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 22: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 23: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 24: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 25: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 26: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 27: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 28: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 29: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 30: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 31: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: CPU KV buffer size = 8.00 MiB
llama_new_context_with_model: KV self size = 8.00 MiB, K (f16): 4.00 MiB, V (f16): 4.00 MiB
llama_new_context_with_model: CPU output buffer size = 0.20 MiB
llama_new_context_with_model: CPU compute buffer size = 6.76 MiB
llama_new_context_with_model: graph nodes = 1030
llama_new_context_with_model: graph splits = 1
llama_new_context_with_model: n_seq_max = 1
llama_new_context_with_model: n_ctx = 96
llama_new_context_with_model: n_ctx_per_seq = 96
llama_new_context_with_model: n_batch = 36
llama_new_context_with_model: n_ubatch = 36
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_new_context_with_model: n_ctx_per_seq (96) < n_ctx_train (4096) -- the full capacity of the model will not be utilized
llama_kv_cache_init: kv_size = 96, offload = 1, type_k = 'f16', type_v = 'f16', n_layer = 32
llama_kv_cache_init: layer 0: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 1: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 2: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 3: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 4: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 5: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 6: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 7: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 8: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 9: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 10: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 11: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 12: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 13: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 14: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 15: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 16: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 17: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 18: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 19: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 20: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 21: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 22: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 23: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 24: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 25: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 26: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 27: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 28: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 29: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 30: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 31: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: CPU KV buffer size = 12.00 MiB
llama_new_context_with_model: KV self size = 12.00 MiB, K (f16): 6.00 MiB, V (f16): 6.00 MiB
llama_new_context_with_model: CPU output buffer size = 0.20 MiB
llama_new_context_with_model: CPU compute buffer size = 7.60 MiB
llama_new_context_with_model: graph nodes = 1030
llama_new_context_with_model: graph splits = 1
llama_new_context_with_model: n_seq_max = 1
llama_new_context_with_model: n_ctx = 128
llama_new_context_with_model: n_ctx_per_seq = 128
llama_new_context_with_model: n_batch = 83
llama_new_context_with_model: n_ubatch = 83
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_new_context_with_model: n_ctx_per_seq (128) < n_ctx_train (4096) -- the full capacity of the model will not be utilized
llama_kv_cache_init: kv_size = 128, offload = 1, type_k = 'f16', type_v = 'f16', n_layer = 32
llama_kv_cache_init: layer 0: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 1: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 2: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 3: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 4: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 5: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 6: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 7: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 8: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 9: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 10: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 11: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 12: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 13: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 14: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 15: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 16: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 17: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 18: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 19: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 20: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 21: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 22: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 23: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 24: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 25: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 26: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 27: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 28: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 29: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 30: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 31: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: CPU KV buffer size = 16.00 MiB
llama_new_context_with_model: KV self size = 16.00 MiB, K (f16): 8.00 MiB, V (f16): 8.00 MiB
llama_new_context_with_model: CPU output buffer size = 0.20 MiB
llama_new_context_with_model: CPU compute buffer size = 17.53 MiB
llama_new_context_with_model: graph nodes = 1030
llama_new_context_with_model: graph splits = 1
llama_new_context_with_model: n_batch is less than GGML_KQ_MASK_PAD - increasing to 32
llama_new_context_with_model: n_seq_max = 1
llama_new_context_with_model: n_ctx = 64
llama_new_context_with_model: n_ctx_per_seq = 64
llama_new_context_with_model: n_batch = 32
llama_new_context_with_model: n_ubatch = 32
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_new_context_with_model: n_ctx_per_seq (64) < n_ctx_train (4096) -- the full capacity of the model will not be utilized
llama_kv_cache_init: kv_size = 64, offload = 1, type_k = 'f16', type_v = 'f16', n_layer = 32
llama_kv_cache_init: layer 0: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 1: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 2: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 3: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 4: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 5: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 6: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 7: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 8: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 9: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 10: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 11: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 12: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 13: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 14: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 15: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 16: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 17: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 18: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 19: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 20: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 21: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 22: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 23: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 24: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 25: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 26: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 27: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 28: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 29: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 30: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 31: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: CPU KV buffer size = 8.00 MiB
llama_new_context_with_model: KV self size = 8.00 MiB, K (f16): 4.00 MiB, V (f16): 4.00 MiB
llama_new_context_with_model: CPU output buffer size = 0.20 MiB
llama_new_context_with_model: CPU compute buffer size = 6.76 MiB
llama_new_context_with_model: graph nodes = 1030
llama_new_context_with_model: graph splits = 1
llama_new_context_with_model: n_seq_max = 1
llama_new_context_with_model: n_ctx = 96
llama_new_context_with_model: n_ctx_per_seq = 96
llama_new_context_with_model: n_batch = 44
llama_new_context_with_model: n_ubatch = 44
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_new_context_with_model: n_ctx_per_seq (96) < n_ctx_train (4096) -- the full capacity of the model will not be utilized
llama_kv_cache_init: kv_size = 96, offload = 1, type_k = 'f16', type_v = 'f16', n_layer = 32
llama_kv_cache_init: layer 0: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 1: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 2: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 3: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 4: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 5: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 6: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 7: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 8: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 9: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 10: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 11: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 12: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 13: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 14: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 15: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 16: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 17: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 18: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 19: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 20: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 21: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 22: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 23: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 24: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 25: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 26: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 27: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 28: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 29: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 30: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: layer 31: n_embd_k_gqa = 1024, n_embd_v_gqa = 1024
llama_kv_cache_init: CPU KV buffer size = 12.00 MiB
llama_new_context_with_model: KV self size = 12.00 MiB, K (f16): 6.00 MiB, V (f16): 6.00 MiB
llama_new_context_with_model: CPU output buffer size = 0.20 MiB
llama_new_context_with_model: CPU compute buffer size = 9.29 MiB
llama_new_context_with_model: graph nodes = 1030
llama_new_context_with_model: graph splits = 1
Processed 5 items
Tokens per second: 1.45
2025-02-15 01:27:09 (262961): called boinc_finish(0)
</stderr_txt>
]]>
©2025 Matteo Rinaldi