llama_model_loader: loaded meta data with 34 key-value pairs and 795 tensors from Mistral-Large-Instruct-2407-IMat-GGUF/Mistral-Large-Instruct-2407.Q8_0.gguf.hardlink.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.name str = Mistral Large Instruct 2407 llama_model_loader: - kv 3: general.version str = 2407 llama_model_loader: - kv 4: general.finetune str = Instruct llama_model_loader: - kv 5: general.basename str = Mistral llama_model_loader: - kv 6: general.size_label str = Large llama_model_loader: - kv 7: general.license str = other llama_model_loader: - kv 8: general.license.name str = mrl llama_model_loader: - kv 9: general.license.link str = https://mistral.ai/licenses/MRL-0.1.md llama_model_loader: - kv 10: general.languages arr[str,10] = ["en", "fr", "de", "es", "it", "pt", ... llama_model_loader: - kv 11: llama.block_count u32 = 88 llama_model_loader: - kv 12: llama.context_length u32 = 32768 llama_model_loader: - kv 13: llama.embedding_length u32 = 12288 llama_model_loader: - kv 14: llama.feed_forward_length u32 = 28672 llama_model_loader: - kv 15: llama.attention.head_count u32 = 96 llama_model_loader: - kv 16: llama.attention.head_count_kv u32 = 8 llama_model_loader: - kv 17: llama.rope.freq_base f32 = 1000000.000000 llama_model_loader: - kv 18: llama.attention.layer_norm_rms_epsilon f32 = 0.000010 llama_model_loader: - kv 19: general.file_type u32 = 7 llama_model_loader: - kv 20: llama.vocab_size u32 = 32768 llama_model_loader: - kv 21: llama.rope.dimension_count u32 = 128 llama_model_loader: - kv 22: tokenizer.ggml.add_space_prefix bool = false llama_model_loader: - kv 23: tokenizer.ggml.model str = llama llama_model_loader: - kv 24: tokenizer.ggml.pre str = default llama_model_loader: - kv 25: tokenizer.ggml.tokens arr[str,32768] = ["", "", "", "[INST]", "[... llama_model_loader: - kv 26: tokenizer.ggml.scores arr[f32,32768] = [-1000.000000, -1000.000000, -1000.00... llama_model_loader: - kv 27: tokenizer.ggml.token_type arr[i32,32768] = [3, 3, 3, 4, 4, 4, 4, 4, 4, 4, 4, 4, ... llama_model_loader: - kv 28: tokenizer.ggml.bos_token_id u32 = 1 llama_model_loader: - kv 29: tokenizer.ggml.eos_token_id u32 = 2 llama_model_loader: - kv 30: tokenizer.ggml.unknown_token_id u32 = 0 llama_model_loader: - kv 31: tokenizer.ggml.add_bos_token bool = true llama_model_loader: - kv 32: tokenizer.ggml.add_eos_token bool = false llama_model_loader: - kv 33: general.quantization_version u32 = 2 llama_model_loader: - type f32: 177 tensors llama_model_loader: - type q8_0: 618 tensors llm_load_vocab: special tokens cache size = 771 llm_load_vocab: token to piece cache size = 0.1732 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 = 32768 llm_load_print_meta: n_merges = 0 llm_load_print_meta: vocab_only = 0 llm_load_print_meta: n_ctx_train = 32768 llm_load_print_meta: n_embd = 12288 llm_load_print_meta: n_layer = 88 llm_load_print_meta: n_head = 96 llm_load_print_meta: n_head_kv = 8 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 = 12 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 = 1000000.0 llm_load_print_meta: freq_scale_train = 1 llm_load_print_meta: n_ctx_orig_yarn = 32768 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 = Q8_0 llm_load_print_meta: model params = 122.61 B llm_load_print_meta: model size = 121.33 GiB (8.50 BPW) llm_load_print_meta: general.name = Mistral Large Instruct 2407 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 = 781 '<0x0A>' llm_load_print_meta: max token length = 48 ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no 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 15 repeating layers to GPU llm_load_tensors: offloaded 15/89 layers to GPU llm_load_tensors: CPU buffer size = 124244.30 MiB llm_load_tensors: CUDA0 buffer size = 21038.91 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 = 1000000.0 llama_new_context_with_model: freq_scale = 1 llama_kv_cache_init: CUDA_Host KV buffer size = 146.00 MiB llama_kv_cache_init: CUDA0 KV buffer size = 30.00 MiB llama_new_context_with_model: KV self size = 176.00 MiB, K (f16): 88.00 MiB, V (f16): 88.00 MiB llama_new_context_with_model: CUDA_Host output buffer size = 0.12 MiB llama_new_context_with_model: CUDA0 compute buffer size = 558.00 MiB llama_new_context_with_model: CUDA_Host compute buffer size = 25.01 MiB llama_new_context_with_model: graph nodes = 2822 llama_new_context_with_model: graph splits = 807 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 92.129 ms compute_imatrix: computing over 148 chunks with batch_size 512 compute_imatrix: 10.14 seconds per pass - ETA 25.00 minutes [1]2.7652,[2]2.3036,[3]2.3357,[4]2.4206,[5]2.7255,[6]2.7036,[7]2.5011,[8]2.7587,[9]2.7426, save_imatrix: stored collected data after 10 chunks in Mistral-Large-Instruct-2407-IMat-GGUF/imatrix.dat [10]3.0130,[11]3.1532,[12]2.9855,[13]3.1570,[14]3.3667,[15]3.6348,[16]3.7626,[17]3.9142,[18]4.0078,[19]4.0978, save_imatrix: stored collected data after 20 chunks in Mistral-Large-Instruct-2407-IMat-GGUF/imatrix.dat [20]4.1987,[21]4.1635,[22]4.0218,[23]3.9708,[24]3.9395,[25]3.9206,[26]3.8255,[27]3.9678,[28]3.9319,[29]4.0183, save_imatrix: stored collected data after 30 chunks in Mistral-Large-Instruct-2407-IMat-GGUF/imatrix.dat [30]3.9205,[31]3.8060,[32]3.6700,[33]3.5299,[34]3.5545,[35]3.5338,[36]3.4375,[37]3.3703,[38]3.3224,[39]3.2892, save_imatrix: stored collected data after 40 chunks in Mistral-Large-Instruct-2407-IMat-GGUF/imatrix.dat [40]3.2717,[41]3.2803,[42]3.2507,[43]3.2383,[44]3.2063,[45]3.1991,[46]3.2194,[47]3.2147,[48]3.2851,[49]3.3108, save_imatrix: stored collected data after 50 chunks in Mistral-Large-Instruct-2407-IMat-GGUF/imatrix.dat [50]3.2353,[51]3.1627,[52]3.0940,[53]3.0296,[54]2.9784,[55]2.9765,[56]2.9845,[57]3.0515,[58]3.1074,[59]3.1411, save_imatrix: stored collected data after 60 chunks in Mistral-Large-Instruct-2407-IMat-GGUF/imatrix.dat [60]3.1231,[61]3.1415,[62]3.1678,[63]3.1926,[64]3.2484,[65]3.2723,[66]3.3095,[67]3.3409,[68]3.3764,[69]3.4027, save_imatrix: stored collected data after 70 chunks in Mistral-Large-Instruct-2407-IMat-GGUF/imatrix.dat [70]3.4200,[71]3.3986,[72]3.3853,[73]3.3911,[74]3.4057,[75]3.4399,[76]3.4384,[77]3.4602,[78]3.4748,[79]3.4722, save_imatrix: stored collected data after 80 chunks in Mistral-Large-Instruct-2407-IMat-GGUF/imatrix.dat [80]3.4766,[81]3.4711,[82]3.4831,[83]3.4922,[84]3.4999,[85]3.5141,[86]3.5168,[87]3.5179,[88]3.5189,[89]3.5277, save_imatrix: stored collected data after 90 chunks in Mistral-Large-Instruct-2407-IMat-GGUF/imatrix.dat [90]3.5246,[91]3.5201,[92]3.5148,[93]3.5127,[94]3.5321,[95]3.5503,[96]3.5495,[97]3.5551,[98]3.5523,[99]3.5754, save_imatrix: stored collected data after 100 chunks in Mistral-Large-Instruct-2407-IMat-GGUF/imatrix.dat [100]3.5520,[101]3.5525,[102]3.5434,[103]3.5608,[104]3.5769,[105]3.5775,[106]3.5673,[107]3.5548,[108]3.5431,[109]3.5302, save_imatrix: stored collected data after 110 chunks in Mistral-Large-Instruct-2407-IMat-GGUF/imatrix.dat [110]3.5162,[111]3.5045,[112]3.4942,[113]3.4829,[114]3.4712,[115]3.4558,[116]3.4651,[117]3.4839,[118]3.5203,[119]3.5556, save_imatrix: stored collected data after 120 chunks in Mistral-Large-Instruct-2407-IMat-GGUF/imatrix.dat [120]3.5883,[121]3.6361,[122]3.6780,[123]3.6863,[124]3.6965,[125]3.6783,[126]3.6783,[127]3.6714,[128]3.6675,[129]3.6367, save_imatrix: stored collected data after 130 chunks in Mistral-Large-Instruct-2407-IMat-GGUF/imatrix.dat [130]3.6077,[131]3.6317,[132]3.6533,[133]3.6594,[134]3.6600,[135]3.6732,[136]3.6950,[137]3.7048,[138]3.7178,[139]3.7363, save_imatrix: stored collected data after 140 chunks in Mistral-Large-Instruct-2407-IMat-GGUF/imatrix.dat [140]3.7499,[141]3.7508,[142]3.7441,[143]3.7117,[144]3.6783,[145]3.6515,[146]3.6198,[147]3.5900,[148]3.5608, save_imatrix: stored collected data after 148 chunks in Mistral-Large-Instruct-2407-IMat-GGUF/imatrix.dat llama_print_timings: load time = 53288.22 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 = 1476348.40 ms / 75776 tokens ( 19.48 ms per token, 51.33 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 = 1520302.56 ms / 75777 tokens Final estimate: PPL = 3.5608 +/- 0.03687