main: build = 3051 (5921b8f0) main: built with cc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0 for x86_64-linux-gnu main: seed = 1717148996 llama_model_loader: loaded meta data with 24 key-value pairs and 255 tensors from neo_7b_instruct_v0.1-IMat-GGUF/neo_7b_instruct_v0.1.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 = neo_7b_instruct_v0.1 llama_model_loader: - kv 2: llama.block_count u32 = 28 llama_model_loader: - kv 3: llama.context_length u32 = 8192 llama_model_loader: - kv 4: llama.embedding_length u32 = 3072 llama_model_loader: - kv 5: llama.feed_forward_length u32 = 24576 llama_model_loader: - kv 6: llama.attention.head_count u32 = 16 llama_model_loader: - kv 7: llama.attention.head_count_kv u32 = 16 llama_model_loader: - kv 8: llama.rope.freq_base f32 = 10000.000000 llama_model_loader: - kv 9: llama.attention.layer_norm_rms_epsilon f32 = 0.000010 llama_model_loader: - kv 10: general.file_type u32 = 0 llama_model_loader: - kv 11: llama.vocab_size u32 = 64256 llama_model_loader: - kv 12: llama.rope.dimension_count u32 = 192 llama_model_loader: - kv 13: tokenizer.ggml.model str = llama llama_model_loader: - kv 14: tokenizer.ggml.pre str = default llama_model_loader: - kv 15: tokenizer.ggml.tokens arr[str,64256] = ["", "~~", "~~", "<0x00>", "<... llama_model_loader: - kv 16: tokenizer.ggml.scores arr[f32,64256] = [0.000000, 0.000000, 0.000000, 0.0000... llama_model_loader: - kv 17: tokenizer.ggml.token_type arr[i32,64256] = [2, 3, 3, 6, 6, 6, 6, 6, 6, 6, 6, 6, ... llama_model_loader: - kv 18: tokenizer.ggml.bos_token_id u32 = 1 llama_model_loader: - kv 19: tokenizer.ggml.eos_token_id u32 = 2 llama_model_loader: - kv 20: tokenizer.ggml.add_bos_token bool = false llama_model_loader: - kv 21: tokenizer.ggml.add_eos_token bool = false llama_model_loader: - kv 22: tokenizer.chat_template str = {% set system_message = 'You are a he... llama_model_loader: - kv 23: general.quantization_version u32 = 2 llama_model_loader: - type f32: 255 tensors llm_load_vocab: special tokens cache size = 515 llm_load_vocab: token to piece cache size = 0.7263 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 = 64256 llm_load_print_meta: n_merges = 0 llm_load_print_meta: n_ctx_train = 8192 llm_load_print_meta: n_embd = 3072 llm_load_print_meta: n_head = 16 llm_load_print_meta: n_head_kv = 16 llm_load_print_meta: n_layer = 28 llm_load_print_meta: n_rot = 192 llm_load_print_meta: n_embd_head_k = 192 llm_load_print_meta: n_embd_head_v = 192 llm_load_print_meta: n_gqa = 1 llm_load_print_meta: n_embd_k_gqa = 3072 llm_load_print_meta: n_embd_v_gqa = 3072 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 = 24576 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_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 = ?B llm_load_print_meta: model ftype = all F32 llm_load_print_meta: model params = 7.79 B llm_load_print_meta: model size = 29.03 GiB (32.00 BPW) llm_load_print_meta: general.name = neo_7b_instruct_v0.1 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 = 13 '<0x0A>' 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 = 0.26 MiB llm_load_tensors: offloading 22 repeating layers to GPU llm_load_tensors: offloaded 22/29 layers to GPU llm_load_tensors: CPU buffer size = 29730.67 MiB llm_load_tensors: CUDA0 buffer size = 22176.52 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 = 10000.0 llama_new_context_with_model: freq_scale = 1 llama_kv_cache_init: CUDA_Host KV buffer size = 36.00 MiB llama_kv_cache_init: CUDA0 KV buffer size = 132.00 MiB llama_new_context_with_model: KV self size = 168.00 MiB, K (f16): 84.00 MiB, V (f16): 84.00 MiB llama_new_context_with_model: CUDA_Host output buffer size = 0.25 MiB llama_new_context_with_model: CUDA0 compute buffer size = 884.50 MiB llama_new_context_with_model: CUDA_Host compute buffer size = 13.01 MiB llama_new_context_with_model: graph nodes = 902 llama_new_context_with_model: graph splits = 70 system_info: n_threads = 25 / 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 100.831 ms compute_imatrix: computing over 149 chunks with batch_size 512 compute_imatrix: 0.88 seconds per pass - ETA 2.18 minutes [1]7.4928,[2]5.5084,[3]5.9895,[4]6.0766,[5]5.8353,[6]5.8586,[7]5.0467,[8]5.1681,[9]5.2769, save_imatrix: stored collected data after 10 chunks in neo_7b_instruct_v0.1-IMat-GGUF/imatrix.dat [10]5.7095,[11]5.6574,[12]5.2763,[13]5.6425,[14]6.1913,[15]6.4770,[16]7.0144,[17]7.2911,[18]7.6367,[19]7.8184, save_imatrix: stored collected data after 20 chunks in neo_7b_instruct_v0.1-IMat-GGUF/imatrix.dat [20]8.2666,[21]7.9858,[22]7.9910,[23]8.0845,[24]8.2336,[25]8.1737,[26]8.1559,[27]8.3450,[28]8.7071,[29]8.9219, save_imatrix: stored collected data after 30 chunks in neo_7b_instruct_v0.1-IMat-GGUF/imatrix.dat [30]8.9232,[31]9.0152,[32]9.0914,[33]9.1701,[34]9.0581,[35]8.9510,[36]8.5395,[37]8.3038,[38]7.9144,[39]7.8461, save_imatrix: stored collected data after 40 chunks in neo_7b_instruct_v0.1-IMat-GGUF/imatrix.dat [40]7.7504,[41]7.6792,[42]7.6430,[43]7.6738,[44]7.6665,[45]7.5552,[46]7.6461,[47]7.7527,[48]7.9371,[49]7.9609, save_imatrix: stored collected data after 50 chunks in neo_7b_instruct_v0.1-IMat-GGUF/imatrix.dat [50]8.1817,[51]8.3743,[52]8.6016,[53]8.7995,[54]8.9741,[55]8.8877,[56]8.9159,[57]9.0592,[58]9.1610,[59]9.1259, save_imatrix: stored collected data after 60 chunks in neo_7b_instruct_v0.1-IMat-GGUF/imatrix.dat [60]9.0814,[61]9.1560,[62]9.2391,[63]9.3159,[64]9.4822,[65]9.5550,[66]9.6800,[67]9.6996,[68]9.7757,[69]9.8577, save_imatrix: stored collected data after 70 chunks in neo_7b_instruct_v0.1-IMat-GGUF/imatrix.dat [70]9.8398,[71]9.7097,[72]9.5802,[73]9.5659,[74]9.6717,[75]9.7145,[76]9.6456,[77]9.6004,[78]9.5912,[79]9.5281, save_imatrix: stored collected data after 80 chunks in neo_7b_instruct_v0.1-IMat-GGUF/imatrix.dat [80]9.4125,[81]9.1966,[82]9.0072,[83]9.0005,[84]9.0031,[85]8.9982,[86]8.9329,[87]8.9485,[88]8.9323,[89]8.9141, save_imatrix: stored collected data after 90 chunks in neo_7b_instruct_v0.1-IMat-GGUF/imatrix.dat [90]8.8906,[91]8.8515,[92]8.8344,[93]8.7844,[94]8.6800,[95]8.7282,[96]8.7940,[97]8.8092,[98]8.7640,[99]8.7790, save_imatrix: stored collected data after 100 chunks in neo_7b_instruct_v0.1-IMat-GGUF/imatrix.dat [100]8.8118,[101]8.7279,[102]8.6677,[103]8.6548,[104]8.6927,[105]8.7042,[106]8.7628,[107]8.8228,[108]8.7613,[109]8.6949, save_imatrix: stored collected data after 110 chunks in neo_7b_instruct_v0.1-IMat-GGUF/imatrix.dat [110]8.6325,[111]8.5552,[112]8.4754,[113]8.4013,[114]8.3407,[115]8.2839,[116]8.2479,[117]8.2627,[118]8.2651,[119]8.3468, save_imatrix: stored collected data after 120 chunks in neo_7b_instruct_v0.1-IMat-GGUF/imatrix.dat [120]8.4304,[121]8.5025,[122]8.5898,[123]8.6898,[124]8.6734,[125]8.7090,[126]8.7099,[127]8.7464,[128]8.7389,[129]8.7467, save_imatrix: stored collected data after 130 chunks in neo_7b_instruct_v0.1-IMat-GGUF/imatrix.dat [130]8.7163,[131]8.6817,[132]8.7224,[133]8.7873,[134]8.7837,[135]8.7811,[136]8.7941,[137]8.8200,[138]8.8213,[139]8.8316, save_imatrix: stored collected data after 140 chunks in neo_7b_instruct_v0.1-IMat-GGUF/imatrix.dat [140]8.8641,[141]8.9160,[142]8.9037,[143]8.9710,[144]9.0113,[145]9.0443,[146]9.1013,[147]9.1251,[148]9.1768,[149]9.1984, save_imatrix: stored collected data after 149 chunks in neo_7b_instruct_v0.1-IMat-GGUF/imatrix.dat llama_print_timings: load time = 3029.26 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 = 119329.02 ms / 76288 tokens ( 1.56 ms per token, 639.31 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 = 122155.95 ms / 76289 tokens Final estimate: PPL = 9.1984 +/- 0.14201