build: 3906 (7eee341b) with Apple clang version 15.0.0 (clang-1500.3.9.4) for arm64-apple-darwin23.6.0 llama_model_loader: loaded meta data with 35 key-value pairs and 219 tensors from salamandra-2b-instruct_Q3_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 = llama llama_model_loader: - kv 1: general.type str = model llama_model_loader: - kv 2: general.size_label str = 2.3B llama_model_loader: - kv 3: general.license str = apache-2.0 llama_model_loader: - kv 4: general.tags arr[str,1] = ["text-generation"] llama_model_loader: - kv 5: general.languages arr[str,36] = ["bg", "ca", "code", "cs", "cy", "da"... llama_model_loader: - kv 6: llama.block_count u32 = 24 llama_model_loader: - kv 7: llama.context_length u32 = 8192 llama_model_loader: - kv 8: llama.embedding_length u32 = 2048 llama_model_loader: - kv 9: llama.feed_forward_length u32 = 5440 llama_model_loader: - kv 10: llama.attention.head_count u32 = 16 llama_model_loader: - kv 11: llama.attention.head_count_kv u32 = 16 llama_model_loader: - kv 12: llama.rope.freq_base f32 = 10000.000000 llama_model_loader: - kv 13: llama.attention.layer_norm_rms_epsilon f32 = 0.000010 llama_model_loader: - kv 14: general.file_type u32 = 12 llama_model_loader: - kv 15: llama.vocab_size u32 = 256000 llama_model_loader: - kv 16: llama.rope.dimension_count u32 = 128 llama_model_loader: - kv 17: tokenizer.ggml.add_space_prefix bool = true llama_model_loader: - kv 18: tokenizer.ggml.model str = llama llama_model_loader: - kv 19: tokenizer.ggml.pre str = default llama_model_loader: - kv 20: tokenizer.ggml.tokens arr[str,256000] = ["", "~~", "~~", "", "<|... llama_model_loader: - kv 21: tokenizer.ggml.scores arr[f32,256000] = [-1000.000000, -1000.000000, -1000.00... llama_model_loader: - kv 22: tokenizer.ggml.token_type arr[i32,256000] = [3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, ... llama_model_loader: - kv 23: tokenizer.ggml.bos_token_id u32 = 1 llama_model_loader: - kv 24: tokenizer.ggml.eos_token_id u32 = 2 llama_model_loader: - kv 25: tokenizer.ggml.unknown_token_id u32 = 0 llama_model_loader: - kv 26: tokenizer.ggml.padding_token_id u32 = 0 llama_model_loader: - kv 27: tokenizer.ggml.add_bos_token bool = true llama_model_loader: - kv 28: tokenizer.ggml.add_eos_token bool = false llama_model_loader: - kv 29: tokenizer.chat_template str = {%- if not date_string is defined %}{... llama_model_loader: - kv 30: general.quantization_version u32 = 2 llama_model_loader: - kv 31: quantize.imatrix.file str = imatrix/oscar/imatrix.dat llama_model_loader: - kv 32: quantize.imatrix.dataset str = ./imatrix/oscar/imatrix-dataset.txt llama_model_loader: - kv 33: quantize.imatrix.entries_count i32 = 168 llama_model_loader: - kv 34: quantize.imatrix.chunks_count i32 = 44176 llama_model_loader: - type f32: 49 tensors llama_model_loader: - type q5_0: 23 tensors llama_model_loader: - type q5_1: 1 tensors llama_model_loader: - type q3_K: 97 tensors llama_model_loader: - type q4_K: 46 tensors llama_model_loader: - type q5_K: 2 tensors llama_model_loader: - type bf16: 1 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 = 104 llm_load_vocab: token to piece cache size = 1.8842 MB llm_load_print_meta: format = GGUF V3 (latest) llm_load_print_meta: arch = llama 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 = 2048 llm_load_print_meta: n_layer = 24 llm_load_print_meta: n_head = 16 llm_load_print_meta: n_head_kv = 16 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 = 1 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-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 = 5440 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 = 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 = ?B llm_load_print_meta: model ftype = Q3_K - Medium llm_load_print_meta: model params = 2.25 B llm_load_print_meta: model size = 1.76 GiB (6.71 BPW) llm_load_print_meta: general.name = n/a llm_load_print_meta: BOS token = 1 '~~' llm_load_print_meta: EOS token = 2 '~~' llm_load_print_meta: UNK token = 0 '' llm_load_print_meta: PAD token = 0 '' llm_load_print_meta: LF token = 145 '<0x0A>' llm_load_print_meta: EOT token = 5 '<|im_end|>' llm_load_print_meta: EOG token = 2 '' llm_load_print_meta: EOG token = 5 '<|im_end|>' llm_load_print_meta: max token length = 72 llm_load_tensors: ggml ctx size = 0.20 MiB llm_load_tensors: offloading 24 repeating layers to GPU llm_load_tensors: offloading non-repeating layers to GPU llm_load_tensors: offloaded 25/25 layers to GPU llm_load_tensors: Metal buffer size = 1801.85 MiB llm_load_tensors: CPU buffer size = 214.84 MiB ................................... llama_new_context_with_model: n_ctx = 8192 llama_new_context_with_model: n_batch = 512 llama_new_context_with_model: n_ubatch = 128 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 ggml_metal_init: allocating ggml_metal_init: found device: Apple M3 Max ggml_metal_init: picking default device: Apple M3 Max ggml_metal_init: using embedded metal library ggml_metal_init: GPU name: Apple M3 Max ggml_metal_init: GPU family: MTLGPUFamilyApple9 (1009) ggml_metal_init: GPU family: MTLGPUFamilyCommon3 (3003) ggml_metal_init: GPU family: MTLGPUFamilyMetal3 (5001) ggml_metal_init: simdgroup reduction support = true ggml_metal_init: simdgroup matrix mul. support = true ggml_metal_init: hasUnifiedMemory = true ggml_metal_init: recommendedMaxWorkingSetSize = 42949.67 MB llama_kv_cache_init: Metal KV buffer size = 1536.00 MiB llama_new_context_with_model: KV self size = 1536.00 MiB, K (f16): 768.00 MiB, V (f16): 768.00 MiB llama_new_context_with_model: CPU output buffer size = 0.98 MiB llama_new_context_with_model: Metal compute buffer size = 72.00 MiB llama_new_context_with_model: CPU compute buffer size = 125.00 MiB llama_new_context_with_model: graph nodes = 774 llama_new_context_with_model: graph splits = 3 common_init_from_params: warming up the model with an empty run - please wait ... (--no-warmup to disable) system_info: n_threads = 15 (n_threads_batch = 15) / 16 | AVX = 0 | AVX_VNNI = 0 | AVX2 = 0 | AVX512 = 0 | AVX512_VBMI = 0 | AVX512_VNNI = 0 | AVX512_BF16 = 0 | FMA = 0 | NEON = 1 | SVE = 0 | ARM_FMA = 1 | F16C = 0 | FP16_VA = 1 | RISCV_VECT = 0 | WASM_SIMD = 0 | BLAS = 1 | SSE3 = 0 | SSSE3 = 0 | VSX = 0 | MATMUL_INT8 = 1 | LLAMAFILE = 1 | perplexity: tokenizing the input .. perplexity: tokenization took 3245.03 ms perplexity: calculating perplexity over 134 chunks, n_ctx=8192, batch_size=512, n_seq=1 perplexity: 10.50 seconds per pass - ETA 23.45 minutes [1]18.6436,[2]19.1149,[3]17.2307,[4]16.9491,[5]16.1724,[6]15.6363,[7]16.6124,[8]16.0910,[9]15.7972,[10]15.0672,[11]15.7426,[12]15.8047,[13]16.9639,[14]17.2806,[15]17.2621,[16]17.8314,[17]18.1578,[18]18.0537,[19]18.0718,[20]18.4048,[21]18.4235,[22]16.3393,[23]16.5358,[24]16.1249,[25]15.5668,[26]15.0885,[27]14.8773,[28]14.6864,[29]14.6278,[30]14.3999,[31]14.6618,[32]14.7695,[33]15.2614,[34]15.5712,[35]15.8861,[36]15.6201,[37]15.6016,[38]15.6714,[39]15.5003,[40]15.5259,[41]15.5006,[42]15.2864,[43]15.2186,[44]15.3937,[45]15.6112,[46]15.4493,[47]15.7212,[48]15.8468,[49]16.1577,[50]16.4769,[51]16.5221,[52]16.7616,[53]17.1090,[54]17.4590,[55]17.5842,[56]17.4052,[57]17.3113,[58]17.0118,[59]16.8908,[60]16.6834,[61]16.7317,[62]16.8849,[63]17.0935,[64]17.1618,[65]17.1949,[66]17.3988,[67]17.3709,[68]17.2497,[69]17.0926,[70]16.9750,[71]16.9734,[72]16.9132,[73]16.9217,[74]16.8675,[75]16.8623,[76]16.8000,[77]16.8626,[78]16.8608,[79]16.8673,[80]16.9007,[81]16.5635,[82]16.5366,[83]16.3972,[84]16.4370,[85]16.4925,[86]16.7065,[87]16.7367,[88]16.9057,[89]16.9635,[90]17.1000,[91]17.1627,[92]16.9849,[93]17.0537,[94]17.0377,[95]17.1891,[96]17.3957,[97]17.4828,[98]17.5903,[99]17.7496,[100]17.7934,[101]17.8236,[102]17.7789,[103]17.7470,[104]17.7305,[105]17.7120,[106]17.5711,[107]17.4287,[108]17.4930,[109]17.5125,[110]17.4178,[111]17.3773,[112]17.2224,[113]17.0680,[114]17.0578,[115]17.0285,[116]17.0352,[117]16.9202,[118]16.7823,[119]16.7747,[120]16.8409,[121]16.8578,[122]16.8858,[123]16.9270,[124]16.9460,[125]16.9417,[126]16.9712,[127]16.9982,[128]17.0843,[129]17.0735,[130]17.0460,[131]17.1078,[132]17.0812,[133]17.0198,[134]16.8567, Final estimate: PPL = 16.8567 +/- 0.06889 llama_perf_context_print: load time = 1230.31 ms llama_perf_context_print: prompt eval time = 1383629.65 ms / 1097728 tokens ( 1.26 ms per token, 793.37 tokens per second) llama_perf_context_print: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) llama_perf_context_print: total time = 1423701.91 ms / 1097729 tokens ggml_metal_free: deallocating