main: build = 3086 (554c247c) main: built with cc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0 for x86_64-linux-gnu main: seed = 1717675894 llama_model_loader: loaded meta data with 23 key-value pairs and 723 tensors from Higgs-Llama-3-70B-IMat-GGUF/Higgs-Llama-3-70B.Q8_0.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.name str = Higgs-Llama-3-70B llama_model_loader: - kv 2: llama.block_count u32 = 80 llama_model_loader: - kv 3: llama.context_length u32 = 8192 llama_model_loader: - kv 4: llama.embedding_length u32 = 8192 llama_model_loader: - kv 5: llama.feed_forward_length u32 = 28672 llama_model_loader: - kv 6: llama.attention.head_count u32 = 64 llama_model_loader: - kv 7: llama.attention.head_count_kv u32 = 8 llama_model_loader: - kv 8: llama.rope.freq_base f32 = 500000.000000 llama_model_loader: - kv 9: llama.attention.layer_norm_rms_epsilon f32 = 0.000010 llama_model_loader: - kv 10: general.file_type u32 = 7 llama_model_loader: - kv 11: llama.vocab_size u32 = 128256 llama_model_loader: - kv 12: llama.rope.dimension_count u32 = 128 llama_model_loader: - kv 13: tokenizer.ggml.model str = gpt2 llama_model_loader: - kv 14: tokenizer.ggml.pre str = smaug-bpe llama_model_loader: - kv 15: tokenizer.ggml.tokens arr[str,128256] = ["!", "\"", "#", "$", "%", "&", "'", ... llama_model_loader: - kv 16: tokenizer.ggml.token_type arr[i32,128256] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... llama_model_loader: - kv 17: tokenizer.ggml.merges arr[str,280147] = ["Ġ Ġ", "Ġ ĠĠĠ", "ĠĠ ĠĠ", "... llama_model_loader: - kv 18: tokenizer.ggml.bos_token_id u32 = 128000 llama_model_loader: - kv 19: tokenizer.ggml.eos_token_id u32 = 128001 llama_model_loader: - kv 20: tokenizer.ggml.padding_token_id u32 = 128001 llama_model_loader: - kv 21: tokenizer.chat_template str = {% set loop_messages = messages %}{% ... llama_model_loader: - kv 22: general.quantization_version u32 = 2 llama_model_loader: - type f32: 161 tensors llama_model_loader: - type q8_0: 562 tensors llm_load_vocab: special tokens cache size = 256 llm_load_vocab: token to piece cache size = 0.8000 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 = 128256 llm_load_print_meta: n_merges = 280147 llm_load_print_meta: n_ctx_train = 8192 llm_load_print_meta: n_embd = 8192 llm_load_print_meta: n_head = 64 llm_load_print_meta: n_head_kv = 8 llm_load_print_meta: n_layer = 80 llm_load_print_meta: n_rot = 128 llm_load_print_meta: n_embd_head_k = 128 llm_load_print_meta: n_embd_head_v = 128 llm_load_print_meta: n_gqa = 8 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 = 28672 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 = 500000.0 llm_load_print_meta: freq_scale_train = 1 llm_load_print_meta: n_yarn_orig_ctx = 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: model type = 70B llm_load_print_meta: model ftype = Q8_0 llm_load_print_meta: model params = 70.55 B llm_load_print_meta: model size = 69.82 GiB (8.50 BPW) llm_load_print_meta: general.name = Higgs-Llama-3-70B llm_load_print_meta: BOS token = 128000 '<|begin_of_text|>' llm_load_print_meta: EOS token = 128001 '<|end_of_text|>' llm_load_print_meta: PAD token = 128001 '<|end_of_text|>' llm_load_print_meta: LF token = 128 'Ä' llm_load_print_meta: EOT token = 128009 '<|eot_id|>' ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: CUDA_USE_TENSOR_CORES: yes ggml_cuda_init: found 1 CUDA devices: Device 0: NVIDIA GeForce RTX 4090, compute capability 8.9, VMM: yes llm_load_tensors: ggml ctx size = 0.74 MiB llm_load_tensors: offloading 23 repeating layers to GPU llm_load_tensors: offloaded 23/81 layers to GPU llm_load_tensors: CPU buffer size = 71494.28 MiB llm_load_tensors: CUDA0 buffer size = 19942.44 MiB .................................................................................................... llama_new_context_with_model: n_ctx = 512 llama_new_context_with_model: n_batch = 512 llama_new_context_with_model: n_ubatch = 512 llama_new_context_with_model: flash_attn = 0 llama_new_context_with_model: freq_base = 500000.0 llama_new_context_with_model: freq_scale = 1 llama_kv_cache_init: CUDA_Host KV buffer size = 114.00 MiB llama_kv_cache_init: CUDA0 KV buffer size = 46.00 MiB llama_new_context_with_model: KV self size = 160.00 MiB, K (f16): 80.00 MiB, V (f16): 80.00 MiB llama_new_context_with_model: CUDA_Host output buffer size = 0.49 MiB llama_new_context_with_model: CUDA0 compute buffer size = 1331.12 MiB llama_new_context_with_model: CUDA_Host compute buffer size = 17.01 MiB llama_new_context_with_model: graph nodes = 2566 llama_new_context_with_model: graph splits = 631 system_info: n_threads = 32 / 32 | AVX = 1 | AVX_VNNI = 0 | AVX2 = 1 | AVX512 = 1 | AVX512_VBMI = 1 | AVX512_VNNI = 1 | AVX512_BF16 = 1 | FMA = 1 | NEON = 0 | SVE = 0 | ARM_FMA = 0 | F16C = 1 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 1 | SSE3 = 1 | SSSE3 = 1 | VSX = 0 | MATMUL_INT8 = 0 | LLAMAFILE = 1 | compute_imatrix: tokenizing the input .. compute_imatrix: tokenization took 181.501 ms compute_imatrix: computing over 125 chunks with batch_size 512 compute_imatrix: 15.83 seconds per pass - ETA 32.97 minutes [1]4.6746,[2]3.7264,[3]3.3172,[4]4.0074,[5]4.0752,[6]3.4988,[7]3.4849,[8]3.8297,[9]4.0246, save_imatrix: stored collected data after 10 chunks in Higgs-Llama-3-70B-IMat-GGUF/imatrix.dat [10]3.7410,[11]4.1182,[12]4.4985,[13]4.8429,[14]5.1377,[15]5.3348,[16]5.5907,[17]5.7580,[18]5.5967,[19]5.3528, save_imatrix: stored collected data after 20 chunks in Higgs-Llama-3-70B-IMat-GGUF/imatrix.dat [20]5.3698,[21]5.4181,[22]5.4081,[23]5.5465,[24]5.5286,[25]5.8102,[26]5.8008,[27]5.4435,[28]5.1707,[29]5.1692, save_imatrix: stored collected data after 30 chunks in Higgs-Llama-3-70B-IMat-GGUF/imatrix.dat [30]5.1437,[31]4.9295,[32]4.7268,[33]4.6368,[34]4.5639,[35]4.6244,[36]4.6603,[37]4.6356,[38]4.7010,[39]4.8262, save_imatrix: stored collected data after 40 chunks in Higgs-Llama-3-70B-IMat-GGUF/imatrix.dat [40]4.8947,[41]4.7263,[42]4.5693,[43]4.4226,[44]4.2843,[45]4.2261,[46]4.1950,[47]4.2907,[48]4.3776,[49]4.4714, save_imatrix: stored collected data after 50 chunks in Higgs-Llama-3-70B-IMat-GGUF/imatrix.dat [50]4.4302,[51]4.5180,[52]4.5959,[53]4.6672,[54]4.7153,[55]4.7853,[56]4.8274,[57]4.8785,[58]4.9240,[59]4.9343, save_imatrix: stored collected data after 60 chunks in Higgs-Llama-3-70B-IMat-GGUF/imatrix.dat [60]4.9289,[61]4.9437,[62]4.9834,[63]5.0309,[64]4.9949,[65]4.9789,[66]4.9964,[67]4.9923,[68]4.9953,[69]4.9955, save_imatrix: stored collected data after 70 chunks in Higgs-Llama-3-70B-IMat-GGUF/imatrix.dat [70]5.0128,[71]5.0165,[72]5.0321,[73]5.0233,[74]5.0112,[75]5.0205,[76]5.0366,[77]5.0282,[78]5.0190,[79]5.0533, save_imatrix: stored collected data after 80 chunks in Higgs-Llama-3-70B-IMat-GGUF/imatrix.dat [80]5.0795,[81]5.0735,[82]5.0908,[83]5.1158,[84]5.0649,[85]5.0625,[86]5.0650,[87]5.0743,[88]5.1019,[89]5.1233, save_imatrix: stored collected data after 90 chunks in Higgs-Llama-3-70B-IMat-GGUF/imatrix.dat [90]5.0981,[91]5.0728,[92]5.0463,[93]5.0234,[94]4.9954,[95]4.9709,[96]4.9480,[97]4.9650,[98]4.9980,[99]5.0550, save_imatrix: stored collected data after 100 chunks in Higgs-Llama-3-70B-IMat-GGUF/imatrix.dat [100]5.1048,[101]5.1433,[102]5.2264,[103]5.2540,[104]5.2822,[105]5.2387,[106]5.2514,[107]5.2253,[108]5.1687,[109]5.1074, save_imatrix: stored collected data after 110 chunks in Higgs-Llama-3-70B-IMat-GGUF/imatrix.dat [110]5.1347,[111]5.1748,[112]5.1846,[113]5.1979,[114]5.2238,[115]5.2558,[116]5.2696,[117]5.2861,[118]5.3158,[119]5.2941, save_imatrix: stored collected data after 120 chunks in Higgs-Llama-3-70B-IMat-GGUF/imatrix.dat [120]5.2402,[121]5.1726,[122]5.1055,[123]5.0469,[124]4.9872,[125]4.9270, save_imatrix: stored collected data after 125 chunks in Higgs-Llama-3-70B-IMat-GGUF/imatrix.dat llama_print_timings: load time = 19623.38 ms llama_print_timings: sample time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) llama_print_timings: prompt eval time = 1807072.34 ms / 64000 tokens ( 28.24 ms per token, 35.42 tokens per second) llama_print_timings: eval time = 0.00 ms / 1 runs ( 0.00 ms per token, inf tokens per second) llama_print_timings: total time = 1812278.76 ms / 64001 tokens Final estimate: PPL = 4.9270 +/- 0.06224