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 = 1717122313 llama_model_loader: loaded meta data with 25 key-value pairs and 723 tensors from K2-Chat-IMat-GGUF/K2-Chat.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 = K2-Chat 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 = 22016 llama_model_loader: - kv 6: llama.attention.head_count u32 = 64 llama_model_loader: - kv 7: llama.attention.head_count_kv u32 = 64 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 = 1 llama_model_loader: - kv 11: llama.vocab_size u32 = 32032 llama_model_loader: - kv 12: llama.rope.dimension_count u32 = 128 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,32032] = ["", "~~", "~~", "<0x00>", "<... llama_model_loader: - kv 16: tokenizer.ggml.scores arr[f32,32032] = [0.000000, 0.000000, 0.000000, 0.0000... llama_model_loader: - kv 17: tokenizer.ggml.token_type arr[i32,32032] = [2, 3, 3, 6, 6, 6, 6, 6, 6, 6, 6, 6, ... llama_model_loader: - kv 18: tokenizer.ggml.bos_token_id u32 = 2 llama_model_loader: - kv 19: tokenizer.ggml.eos_token_id u32 = 2 llama_model_loader: - kv 20: tokenizer.ggml.unknown_token_id u32 = 0 llama_model_loader: - kv 21: tokenizer.ggml.padding_token_id u32 = 0 llama_model_loader: - kv 22: tokenizer.ggml.add_bos_token bool = false llama_model_loader: - kv 23: tokenizer.ggml.add_eos_token bool = false llama_model_loader: - kv 24: general.quantization_version u32 = 2 llama_model_loader: - type f32: 161 tensors llama_model_loader: - type f16: 562 tensors llm_load_vocab: special tokens cache size = 291 llm_load_vocab: token to piece cache size = 0.3374 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 = 32032 llm_load_print_meta: n_merges = 0 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 = 64 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 = 1 llm_load_print_meta: n_embd_k_gqa = 8192 llm_load_print_meta: n_embd_v_gqa = 8192 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 = 22016 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 = 65B llm_load_print_meta: model ftype = F16 llm_load_print_meta: model params = 65.29 B llm_load_print_meta: model size = 121.61 GiB (16.00 BPW) llm_load_print_meta: general.name = K2-Chat llm_load_print_meta: BOS token = 2 '' 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 = 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 = 0.74 MiB llm_load_tensors: offloading 14 repeating layers to GPU llm_load_tensors: offloaded 14/81 layers to GPU llm_load_tensors: CPU buffer size = 124526.03 MiB llm_load_tensors: CUDA0 buffer size = 21616.88 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 = 1056.00 MiB llama_kv_cache_init: CUDA0 KV buffer size = 224.00 MiB llama_new_context_with_model: KV self size = 1280.00 MiB, K (f16): 640.00 MiB, V (f16): 640.00 MiB llama_new_context_with_model: CUDA_Host output buffer size = 0.12 MiB llama_new_context_with_model: CUDA0 compute buffer size = 579.06 MiB llama_new_context_with_model: CUDA_Host compute buffer size = 33.01 MiB llama_new_context_with_model: graph nodes = 2566 llama_new_context_with_model: graph splits = 730 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 89.427 ms compute_imatrix: computing over 151 chunks with batch_size 512 compute_imatrix: 102.37 seconds per pass - ETA 4 hours 17.62 minutes [1]2.8917,[2]2.3733,[3]2.4306,[4]2.5238,[5]2.9298,[6]2.9024,[7]2.7012,[8]3.0878,[9]3.2174, save_imatrix: stored collected data after 10 chunks in K2-Chat-IMat-GGUF/imatrix.dat [10]3.4833,[11]3.5745,[12]3.3664,[13]3.4248,[14]3.6486,[15]3.9714,[16]4.1241,[17]4.3969,[18]4.5736,[19]4.7304, save_imatrix: stored collected data after 20 chunks in K2-Chat-IMat-GGUF/imatrix.dat [20]4.8577,[21]5.0109,[22]4.8424,[23]4.6897,[24]4.7672,[25]4.8186,[26]4.8201,[27]4.8100,[28]4.8872,[29]5.0006, save_imatrix: stored collected data after 30 chunks in K2-Chat-IMat-GGUF/imatrix.dat [30]5.1357,[31]5.0993,[32]4.9583,[33]4.8463,[34]4.8234,[35]4.8521,[36]4.8197,[37]4.6568,[38]4.5458,[39]4.5119, save_imatrix: stored collected data after 40 chunks in K2-Chat-IMat-GGUF/imatrix.dat [40]4.4825,[41]4.4482,[42]4.4773,[43]4.4669,[44]4.5030,[45]4.5069,[46]4.5002,[47]4.5301,[48]4.5963,[49]4.6671, save_imatrix: stored collected data after 50 chunks in K2-Chat-IMat-GGUF/imatrix.dat [50]4.6928,[51]4.7477,[52]4.8170,[53]4.9125,[54]4.9989,[55]5.0391,[56]5.0108,[57]4.9710,[58]5.0029,[59]5.0480, save_imatrix: stored collected data after 60 chunks in K2-Chat-IMat-GGUF/imatrix.dat [60]5.1174,[61]5.0568,[62]5.0936,[63]5.1241,[64]5.1943,[65]5.2544,[66]5.2831,[67]5.3325,[68]5.3753,[69]5.3872, save_imatrix: stored collected data after 70 chunks in K2-Chat-IMat-GGUF/imatrix.dat [70]5.3959,[71]5.4032,[72]5.4125,[73]5.3823,[74]5.3346,[75]5.3348,[76]5.3345,[77]5.3445,[78]5.3154,[79]5.3184, save_imatrix: stored collected data after 80 chunks in K2-Chat-IMat-GGUF/imatrix.dat [80]5.3246,[81]5.3009,[82]5.2858,[83]5.2587,[84]5.2593,[85]5.2517,[86]5.2436,[87]5.2418,[88]5.2413,[89]5.2174, save_imatrix: stored collected data after 90 chunks in K2-Chat-IMat-GGUF/imatrix.dat [90]5.2034,[91]5.2041,[92]5.1831,[93]5.1593,[94]5.1425,[95]5.0979,[96]5.1113,[97]5.1006,[98]5.0973,[99]5.0798, save_imatrix: stored collected data after 100 chunks in K2-Chat-IMat-GGUF/imatrix.dat [100]5.0671,[101]5.0824,[102]5.0565,[103]5.0375,[104]5.0312,[105]5.0472,[106]5.0485,[107]5.0669,[108]5.0768,[109]5.0515, save_imatrix: stored collected data after 110 chunks in K2-Chat-IMat-GGUF/imatrix.dat [110]5.0264,[111]5.0004,[112]4.9732,[113]4.9441,[114]4.9172,[115]4.8935,[116]4.8705,[117]4.8572,[118]4.8748,[119]4.8912, save_imatrix: stored collected data after 120 chunks in K2-Chat-IMat-GGUF/imatrix.dat [120]4.9312,[121]4.9665,[122]5.0091,[123]5.0480,[124]5.1134,[125]5.1657,[126]5.1845,[127]5.1971,[128]5.1574,[129]5.1573, save_imatrix: stored collected data after 130 chunks in K2-Chat-IMat-GGUF/imatrix.dat [130]5.1479,[131]5.1375,[132]5.1209,[133]5.1026,[134]5.1199,[135]5.1425,[136]5.1467,[137]5.1523,[138]5.1644,[139]5.1807, save_imatrix: stored collected data after 140 chunks in K2-Chat-IMat-GGUF/imatrix.dat [140]5.1983,[141]5.2080,[142]5.2142,[143]5.2108,[144]5.2018,[145]5.2090,[146]5.1976,[147]5.1849,[148]5.1931,[149]5.1812, save_imatrix: stored collected data after 150 chunks in K2-Chat-IMat-GGUF/imatrix.dat [150]5.1743,[151]5.1552, save_imatrix: stored collected data after 151 chunks in K2-Chat-IMat-GGUF/imatrix.dat llama_print_timings: load time = 209713.66 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 = 1236070.62 ms / 77312 tokens ( 15.99 ms per token, 62.55 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 = 1346358.00 ms / 77313 tokens Final estimate: PPL = 5.1552 +/- 0.05903