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 33 key-value pairs and 219 tensors from salamandra-2b_Q3_K_S.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 = 11 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.add_bos_token bool = true llama_model_loader: - kv 27: tokenizer.ggml.add_eos_token bool = false llama_model_loader: - kv 28: general.quantization_version u32 = 2 llama_model_loader: - kv 29: quantize.imatrix.file str = imatrix/oscar/imatrix.dat llama_model_loader: - kv 30: quantize.imatrix.dataset str = ./imatrix/oscar/imatrix-dataset.txt llama_model_loader: - kv 31: quantize.imatrix.entries_count i32 = 168 llama_model_loader: - kv 32: quantize.imatrix.chunks_count i32 = 44176 llama_model_loader: - type f32: 49 tensors llama_model_loader: - type q3_K: 145 tensors llama_model_loader: - type iq4_nl: 24 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 - Small llm_load_print_meta: model params = 2.25 B llm_load_print_meta: model size = 1.70 GiB (6.49 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: 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 = 1742.81 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 2659.3 ms perplexity: calculating perplexity over 134 chunks, n_ctx=8192, batch_size=512, n_seq=1 perplexity: 10.57 seconds per pass - ETA 23.60 minutes [1]17.7009,[2]17.7183,[3]16.0449,[4]15.8021,[5]15.0242,[6]14.5446,[7]15.3865,[8]14.8945,[9]14.6139,[10]13.9402,[11]14.6111,[12]14.7133,[13]15.7806,[14]16.0797,[15]16.0916,[16]16.6376,[17]16.9555,[18]16.8784,[19]16.9137,[20]17.2161,[21]17.2339,[22]15.3506,[23]15.5269,[24]15.1380,[25]14.6262,[26]14.1965,[27]14.0100,[28]13.8386,[29]13.7831,[30]13.5691,[31]13.8025,[32]13.9226,[33]14.3713,[34]14.6669,[35]14.9580,[36]14.7204,[37]14.6991,[38]14.7626,[39]14.6053,[40]14.6213,[41]14.6021,[42]14.4082,[43]14.3441,[44]14.5114,[45]14.7136,[46]14.5605,[47]14.8247,[48]14.9493,[49]15.2499,[50]15.5027,[51]15.5399,[52]15.7576,[53]16.0764,[54]16.3984,[55]16.5178,[56]16.3493,[57]16.2583,[58]15.9822,[59]15.8691,[60]15.6761,[61]15.7143,[62]15.8554,[63]16.0455,[64]16.0997,[65]16.1167,[66]16.3043,[67]16.2840,[68]16.1785,[69]16.0378,[70]15.9344,[71]15.9397,[72]15.8860,[73]15.8930,[74]15.8387,[75]15.8212,[76]15.7590,[77]15.8149,[78]15.8088,[79]15.8166,[80]15.8450,[81]15.5696,[82]15.5457,[83]15.4155,[84]15.4556,[85]15.5065,[86]15.7108,[87]15.7445,[88]15.9013,[89]15.9568,[90]16.0805,[91]16.1409,[92]15.9790,[93]16.0429,[94]16.0270,[95]16.1650,[96]16.3572,[97]16.4344,[98]16.5345,[99]16.6520,[100]16.6945,[101]16.7270,[102]16.6893,[103]16.6560,[104]16.6379,[105]16.6156,[106]16.4825,[107]16.3513,[108]16.4158,[109]16.4384,[110]16.3460,[111]16.3085,[112]16.1596,[113]16.0174,[114]16.0084,[115]15.9790,[116]15.9850,[117]15.8774,[118]15.7481,[119]15.7459,[120]15.8063,[121]15.8262,[122]15.8534,[123]15.8950,[124]15.9130,[125]15.9119,[126]15.9398,[127]15.9660,[128]16.0439,[129]16.0354,[130]16.0105,[131]16.0663,[132]16.0421,[133]15.9897,[134]15.8390, Final estimate: PPL = 15.8390 +/- 0.06302 llama_perf_context_print: load time = 1170.09 ms llama_perf_context_print: prompt eval time = 1657555.43 ms / 1097728 tokens ( 1.51 ms per token, 662.26 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 = 1698503.99 ms / 1097729 tokens ggml_metal_free: deallocating