main: build = 2998 (9588f196) main: built with cc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0 for x86_64-linux-gnu main: seed = 1716654535 llama_model_loader: loaded meta data with 23 key-value pairs and 291 tensors from RoLlama2-7b-Chat-IMat-GGUF/RoLlama2-7b-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 = RoLlama2-7b-Chat llama_model_loader: - kv 2: llama.block_count u32 = 32 llama_model_loader: - kv 3: llama.context_length u32 = 4096 llama_model_loader: - kv 4: llama.embedding_length u32 = 4096 llama_model_loader: - kv 5: llama.feed_forward_length u32 = 11008 llama_model_loader: - kv 6: llama.attention.head_count u32 = 32 llama_model_loader: - kv 7: llama.attention.head_count_kv u32 = 32 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 = 32004 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,32004] = ["", "~~", "~~", "<0x00>", "<... llama_model_loader: - kv 16: tokenizer.ggml.scores arr[f32,32004] = [0.000000, 0.000000, 0.000000, 0.0000... llama_model_loader: - kv 17: tokenizer.ggml.token_type arr[i32,32004] = [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 = true llama_model_loader: - kv 21: tokenizer.ggml.add_eos_token bool = false llama_model_loader: - kv 22: general.quantization_version u32 = 2 llama_model_loader: - type f32: 291 tensors llm_load_vocab: special tokens definition check successful ( 263/32004 ). 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 = 32004 llm_load_print_meta: n_merges = 0 llm_load_print_meta: n_ctx_train = 4096 llm_load_print_meta: n_embd = 4096 llm_load_print_meta: n_head = 32 llm_load_print_meta: n_head_kv = 32 llm_load_print_meta: n_layer = 32 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 = 4096 llm_load_print_meta: n_embd_v_gqa = 4096 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 = 11008 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 = 4096 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 = 7B llm_load_print_meta: model ftype = all F32 llm_load_print_meta: model params = 6.74 B llm_load_print_meta: model size = 25.10 GiB (32.00 BPW) llm_load_print_meta: general.name = RoLlama2-7b-Chat 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 = 0.30 MiB llm_load_tensors: offloading 20 repeating layers to GPU llm_load_tensors: offloaded 20/33 layers to GPU llm_load_tensors: CPU buffer size = 25705.14 MiB llm_load_tensors: CUDA0 buffer size = 15440.62 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 = 96.00 MiB llama_kv_cache_init: CUDA0 KV buffer size = 160.00 MiB llama_new_context_with_model: KV self size = 256.00 MiB, K (f16): 128.00 MiB, V (f16): 128.00 MiB llama_new_context_with_model: CUDA_Host output buffer size = 0.12 MiB llama_new_context_with_model: CUDA0 compute buffer size = 570.57 MiB llama_new_context_with_model: CUDA_Host compute buffer size = 17.01 MiB llama_new_context_with_model: graph nodes = 1030 llama_new_context_with_model: graph splits = 136 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 132.228 ms compute_imatrix: computing over 234 chunks with batch_size 512 compute_imatrix: 1.05 seconds per pass - ETA 4.08 minutes [1]5.9883,[2]4.3028,[3]4.2195,[4]4.7777,[5]5.3795,[6]5.4479,[7]4.9412,[8]5.3200,[9]5.5114, save_imatrix: stored collected data after 10 chunks in RoLlama2-7b-Chat-IMat-GGUF/imatrix.dat [10]5.8365,[11]5.8661,[12]5.3748,[13]5.5065,[14]5.4479,[15]5.8823,[16]6.0444,[17]6.3551,[18]6.5563,[19]6.7513, save_imatrix: stored collected data after 20 chunks in RoLlama2-7b-Chat-IMat-GGUF/imatrix.dat [20]6.9175,[21]6.9677,[22]7.0994,[23]6.7775,[24]6.5802,[25]6.6010,[26]6.3218,[27]6.1921,[28]5.9742,[29]5.9706, save_imatrix: stored collected data after 30 chunks in RoLlama2-7b-Chat-IMat-GGUF/imatrix.dat [30]6.0998,[31]6.1897,[32]6.2482,[33]6.2075,[34]6.2231,[35]6.2236,[36]5.9746,[37]5.8249,[38]5.7662,[39]5.7353, save_imatrix: stored collected data after 40 chunks in RoLlama2-7b-Chat-IMat-GGUF/imatrix.dat [40]5.6991,[41]5.6180,[42]5.6725,[43]5.6992,[44]5.7500,[45]5.8197,[46]5.8982,[47]5.9433,[48]6.0756,[49]6.1833, save_imatrix: stored collected data after 50 chunks in RoLlama2-7b-Chat-IMat-GGUF/imatrix.dat [50]6.3015,[51]6.3938,[52]6.5002,[53]6.4863,[54]6.4201,[55]6.3737,[56]6.4561,[57]6.5051,[58]6.5286,[59]6.5952, save_imatrix: stored collected data after 60 chunks in RoLlama2-7b-Chat-IMat-GGUF/imatrix.dat [60]6.6776,[61]6.7103,[62]6.7716,[63]6.8134,[64]6.8793,[65]6.8970,[66]6.9295,[67]6.9635,[68]6.9962,[69]7.0409, save_imatrix: stored collected data after 70 chunks in RoLlama2-7b-Chat-IMat-GGUF/imatrix.dat [70]7.0722,[71]7.1080,[72]7.1553,[73]7.1003,[74]7.0536,[75]7.0109,[76]6.9812,[77]6.9803,[78]6.9473,[79]6.9050, save_imatrix: stored collected data after 80 chunks in RoLlama2-7b-Chat-IMat-GGUF/imatrix.dat [80]6.8480,[81]6.8371,[82]6.7966,[83]6.7611,[84]6.7787,[85]6.8018,[86]6.8150,[87]6.8440,[88]6.8422,[89]6.8132, save_imatrix: stored collected data after 90 chunks in RoLlama2-7b-Chat-IMat-GGUF/imatrix.dat [90]6.8072,[91]6.8022,[92]6.7869,[93]6.7982,[94]6.7808,[95]6.7852,[96]6.8023,[97]6.8192,[98]6.7921,[99]6.7531, save_imatrix: stored collected data after 100 chunks in RoLlama2-7b-Chat-IMat-GGUF/imatrix.dat [100]6.7634,[101]6.7825,[102]6.7751,[103]6.7561,[104]6.7218,[105]6.7086,[106]6.7279,[107]6.7382,[108]6.7262,[109]6.7229, save_imatrix: stored collected data after 110 chunks in RoLlama2-7b-Chat-IMat-GGUF/imatrix.dat [110]6.7033,[111]6.7208,[112]6.7393,[113]6.7427,[114]6.7666,[115]6.7688,[116]6.7650,[117]6.7618,[118]6.7739,[119]6.7529, save_imatrix: stored collected data after 120 chunks in RoLlama2-7b-Chat-IMat-GGUF/imatrix.dat [120]6.7519,[121]6.7428,[122]6.7174,[123]6.7380,[124]6.7345,[125]6.7393,[126]6.7300,[127]6.7290,[128]6.7449,[129]6.7188, save_imatrix: stored collected data after 130 chunks in RoLlama2-7b-Chat-IMat-GGUF/imatrix.dat [130]6.6931,[131]6.6752,[132]6.6763,[133]6.6400,[134]6.6423,[135]6.6107,[136]6.5848,[137]6.5509,[138]6.5181,[139]6.4871, save_imatrix: stored collected data after 140 chunks in RoLlama2-7b-Chat-IMat-GGUF/imatrix.dat [140]6.4590,[141]6.4332,[142]6.4075,[143]6.4080,[144]6.4042,[145]6.3873,[146]6.3619,[147]6.3590,[148]6.3526,[149]6.3469, save_imatrix: stored collected data after 150 chunks in RoLlama2-7b-Chat-IMat-GGUF/imatrix.dat [150]6.3401,[151]6.3234,[152]6.3164,[153]6.3043,[154]6.2931,[155]6.3051,[156]6.2827,[157]6.2784,[158]6.2902,[159]6.2846, save_imatrix: stored collected data after 160 chunks in RoLlama2-7b-Chat-IMat-GGUF/imatrix.dat [160]6.2848,[161]6.2951,[162]6.2968,[163]6.3100,[164]6.3169,[165]6.3329,[166]6.3391,[167]6.3384,[168]6.3446,[169]6.3505, save_imatrix: stored collected data after 170 chunks in RoLlama2-7b-Chat-IMat-GGUF/imatrix.dat [170]6.3742,[171]6.3733,[172]6.3863,[173]6.4140,[174]6.4263,[175]6.4508,[176]6.4671,[177]6.4877,[178]6.5032,[179]6.5302, save_imatrix: stored collected data after 180 chunks in RoLlama2-7b-Chat-IMat-GGUF/imatrix.dat [180]6.5455,[181]6.5936,[182]6.6102,[183]6.6312,[184]6.6330,[185]6.6379,[186]6.6479,[187]6.6518,[188]6.6372,[189]6.6427, save_imatrix: stored collected data after 190 chunks in RoLlama2-7b-Chat-IMat-GGUF/imatrix.dat [190]6.6507,[191]6.6596,[192]6.6670,[193]6.6965,[194]6.6913,[195]6.6699,[196]6.7046,[197]6.7383,[198]6.7669,[199]6.8205, save_imatrix: stored collected data after 200 chunks in RoLlama2-7b-Chat-IMat-GGUF/imatrix.dat [200]6.8660,[201]6.8749,[202]6.8785,[203]6.8443,[204]6.8471,[205]6.8582,[206]6.8843,[207]6.8737,[208]6.8706,[209]6.8729, save_imatrix: stored collected data after 210 chunks in RoLlama2-7b-Chat-IMat-GGUF/imatrix.dat [210]6.8845,[211]6.9005,[212]6.8986,[213]6.8943,[214]6.8991,[215]6.9194,[216]6.9322,[217]6.9379,[218]6.9316,[219]6.9273, save_imatrix: stored collected data after 220 chunks in RoLlama2-7b-Chat-IMat-GGUF/imatrix.dat [220]6.9200,[221]6.9196,[222]6.9186,[223]6.9421,[224]6.9281,[225]6.9293,[226]6.9200,[227]6.9536,[228]6.9927,[229]7.0349, save_imatrix: stored collected data after 230 chunks in RoLlama2-7b-Chat-IMat-GGUF/imatrix.dat [230]7.0738,[231]7.0879,[232]7.0712,[233]7.0533,[234]7.0337, save_imatrix: stored collected data after 234 chunks in RoLlama2-7b-Chat-IMat-GGUF/imatrix.dat llama_print_timings: load time = 2820.15 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 = 232394.01 ms / 119808 tokens ( 1.94 ms per token, 515.54 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 = 235165.91 ms / 119809 tokens Final estimate: PPL = 7.0337 +/- 0.06898