main: build = 3003 (d298382a) main: built with cc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0 for x86_64-linux-gnu main: seed = 1716748134 llama_model_loader: loaded meta data with 24 key-value pairs and 291 tensors from Mistral-7B-v0.3-IMat-GGUF/Mistral-7B-v0.3.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 = Mistral-7B-v0.3 llama_model_loader: - kv 2: llama.block_count u32 = 32 llama_model_loader: - kv 3: llama.context_length u32 = 32768 llama_model_loader: - kv 4: llama.embedding_length u32 = 4096 llama_model_loader: - kv 5: llama.feed_forward_length u32 = 14336 llama_model_loader: - kv 6: llama.attention.head_count u32 = 32 llama_model_loader: - kv 7: llama.attention.head_count_kv u32 = 8 llama_model_loader: - kv 8: llama.rope.freq_base f32 = 1000000.000000 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 = 32768 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,32768] = ["<unk>", "<s>", "</s>", "[INST]", "[... llama_model_loader: - kv 16: tokenizer.ggml.scores arr[f32,32768] = [0.000000, 0.000000, 0.000000, 0.0000... llama_model_loader: - kv 17: tokenizer.ggml.token_type arr[i32,32768] = [2, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, ... 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.unknown_token_id u32 = 0 llama_model_loader: - kv 21: tokenizer.ggml.add_bos_token bool = true llama_model_loader: - kv 22: tokenizer.ggml.add_eos_token bool = false llama_model_loader: - kv 23: general.quantization_version u32 = 2 llama_model_loader: - type f32: 291 tensors llm_load_vocab: special tokens definition check successful ( 1027/32768 ). 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: n_ctx_train = 32768 llm_load_print_meta: n_embd = 4096 llm_load_print_meta: n_head = 32 llm_load_print_meta: n_head_kv = 8 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 = 4 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 = 14336 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_yarn_orig_ctx = 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 = 7B llm_load_print_meta: model ftype = all F32 llm_load_print_meta: model params = 7.25 B llm_load_print_meta: model size = 27.00 GiB (32.00 BPW) llm_load_print_meta: general.name = Mistral-7B-v0.3 llm_load_print_meta: BOS token = 1 '<s>' llm_load_print_meta: EOS token = 2 '</s>' llm_load_print_meta: UNK token = 0 '<unk>' llm_load_print_meta: LF token = 781 '<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 25 repeating layers to GPU llm_load_tensors: offloaded 25/33 layers to GPU llm_load_tensors: CPU buffer size = 27649.02 MiB llm_load_tensors: CUDA0 buffer size = 20800.78 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 = 14.00 MiB llama_kv_cache_init: CUDA0 KV buffer size = 50.00 MiB llama_new_context_with_model: KV self size = 64.00 MiB, K (f16): 32.00 MiB, V (f16): 32.00 MiB llama_new_context_with_model: CUDA_Host output buffer size = 0.12 MiB llama_new_context_with_model: CUDA0 compute buffer size = 584.00 MiB llama_new_context_with_model: CUDA_Host compute buffer size = 9.01 MiB llama_new_context_with_model: graph nodes = 1030 llama_new_context_with_model: graph splits = 81 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 133.793 ms compute_imatrix: computing over 228 chunks with batch_size 512 compute_imatrix: 0.84 seconds per pass - ETA 3.18 minutes [1]3.6887,[2]2.7710,[3]2.8053,[4]2.9263,[5]3.2772,[6]3.2192,[7]2.9588,[8]3.3808,[9]3.5098, save_imatrix: stored collected data after 10 chunks in Mistral-7B-v0.3-IMat-GGUF/imatrix.dat [10]3.8600,[11]4.0156,[12]3.9434,[13]4.1868,[14]3.9891,[15]4.3176,[16]4.4575,[17]4.6694,[18]4.7922,[19]4.9315, save_imatrix: stored collected data after 20 chunks in Mistral-7B-v0.3-IMat-GGUF/imatrix.dat [20]5.0376,[21]5.1603,[22]5.0485,[23]4.8667,[24]4.9439,[25]4.7118,[26]4.5371,[27]4.4214,[28]4.3613,[29]4.3522, save_imatrix: stored collected data after 30 chunks in Mistral-7B-v0.3-IMat-GGUF/imatrix.dat [30]4.4437,[31]4.5603,[32]4.6703,[33]4.6912,[34]4.7585,[35]4.6052,[36]4.5137,[37]4.4660,[38]4.4682,[39]4.4583, save_imatrix: stored collected data after 40 chunks in Mistral-7B-v0.3-IMat-GGUF/imatrix.dat [40]4.4212,[41]4.4608,[42]4.4195,[43]4.4770,[44]4.5632,[45]4.5759,[46]4.6612,[47]4.7868,[48]4.8921,[49]5.0255, save_imatrix: stored collected data after 50 chunks in Mistral-7B-v0.3-IMat-GGUF/imatrix.dat [50]5.1063,[51]5.1274,[52]5.0926,[53]5.0579,[54]4.9702,[55]5.0257,[56]5.0751,[57]5.1272,[58]5.1711,[59]5.1854, save_imatrix: stored collected data after 60 chunks in Mistral-7B-v0.3-IMat-GGUF/imatrix.dat [60]5.2564,[61]5.2934,[62]5.3388,[63]5.3548,[64]5.3775,[65]5.4095,[66]5.4457,[67]5.4879,[68]5.5328,[69]5.5510, save_imatrix: stored collected data after 70 chunks in Mistral-7B-v0.3-IMat-GGUF/imatrix.dat [70]5.5758,[71]5.5393,[72]5.5005,[73]5.4730,[74]5.4483,[75]5.4336,[76]5.4238,[77]5.4001,[78]5.3556,[79]5.3346, save_imatrix: stored collected data after 80 chunks in Mistral-7B-v0.3-IMat-GGUF/imatrix.dat [80]5.3310,[81]5.3055,[82]5.3476,[83]5.3814,[84]5.3983,[85]5.3469,[86]5.3646,[87]5.3315,[88]5.2861,[89]5.2791, save_imatrix: stored collected data after 90 chunks in Mistral-7B-v0.3-IMat-GGUF/imatrix.dat [90]5.2740,[91]5.2863,[92]5.2868,[93]5.3020,[94]5.2958,[95]5.2515,[96]5.2190,[97]5.2123,[98]5.2387,[99]5.2521, save_imatrix: stored collected data after 100 chunks in Mistral-7B-v0.3-IMat-GGUF/imatrix.dat [100]5.2495,[101]5.2227,[102]5.2017,[103]5.2084,[104]5.2047,[105]5.1942,[106]5.1811,[107]5.1837,[108]5.1907,[109]5.2046, save_imatrix: stored collected data after 110 chunks in Mistral-7B-v0.3-IMat-GGUF/imatrix.dat [110]5.1906,[111]5.1971,[112]5.1900,[113]5.1854,[114]5.1776,[115]5.1848,[116]5.1825,[117]5.1759,[118]5.1510,[119]5.1563, save_imatrix: stored collected data after 120 chunks in Mistral-7B-v0.3-IMat-GGUF/imatrix.dat [120]5.1792,[121]5.1875,[122]5.1834,[123]5.1921,[124]5.1992,[125]5.2200,[126]5.1767,[127]5.1732,[128]5.1547,[129]5.1287, save_imatrix: stored collected data after 130 chunks in Mistral-7B-v0.3-IMat-GGUF/imatrix.dat [130]5.1449,[131]5.1199,[132]5.0965,[133]5.0710,[134]5.0473,[135]5.0219,[136]4.9990,[137]4.9776,[138]4.9583,[139]4.9439, save_imatrix: stored collected data after 140 chunks in Mistral-7B-v0.3-IMat-GGUF/imatrix.dat [140]4.9404,[141]4.9330,[142]4.9132,[143]4.9075,[144]4.8956,[145]4.8864,[146]4.8768,[147]4.8661,[148]4.8616,[149]4.8490, save_imatrix: stored collected data after 150 chunks in Mistral-7B-v0.3-IMat-GGUF/imatrix.dat [150]4.8407,[151]4.8543,[152]4.8336,[153]4.8377,[154]4.8609,[155]4.8827,[156]4.8943,[157]4.9074,[158]4.9254,[159]4.9562, save_imatrix: stored collected data after 160 chunks in Mistral-7B-v0.3-IMat-GGUF/imatrix.dat [160]4.9729,[161]4.9883,[162]4.9924,[163]4.9994,[164]5.0147,[165]5.0150,[166]5.0230,[167]5.0376,[168]5.0464,[169]5.0597, save_imatrix: stored collected data after 170 chunks in Mistral-7B-v0.3-IMat-GGUF/imatrix.dat [170]5.0592,[171]5.0757,[172]5.0884,[173]5.0913,[174]5.1014,[175]5.0897,[176]5.1058,[177]5.1118,[178]5.1265,[179]5.1225, save_imatrix: stored collected data after 180 chunks in Mistral-7B-v0.3-IMat-GGUF/imatrix.dat [180]5.1375,[181]5.1385,[182]5.1383,[183]5.1360,[184]5.1356,[185]5.1441,[186]5.1511,[187]5.1701,[188]5.1731,[189]5.1560, save_imatrix: stored collected data after 190 chunks in Mistral-7B-v0.3-IMat-GGUF/imatrix.dat [190]5.1865,[191]5.2168,[192]5.2454,[193]5.2908,[194]5.3229,[195]5.3319,[196]5.3409,[197]5.3252,[198]5.3296,[199]5.3455, save_imatrix: stored collected data after 200 chunks in Mistral-7B-v0.3-IMat-GGUF/imatrix.dat [200]5.3699,[201]5.3676,[202]5.3656,[203]5.3719,[204]5.3840,[205]5.3879,[206]5.3943,[207]5.4031,[208]5.4108,[209]5.4244, save_imatrix: stored collected data after 210 chunks in Mistral-7B-v0.3-IMat-GGUF/imatrix.dat [210]5.4449,[211]5.4363,[212]5.4373,[213]5.4324,[214]5.4289,[215]5.4240,[216]5.4178,[217]5.4146,[218]5.4271,[219]5.4128, save_imatrix: stored collected data after 220 chunks in Mistral-7B-v0.3-IMat-GGUF/imatrix.dat [220]5.4227,[221]5.4519,[222]5.4685,[223]5.4930,[224]5.5081,[225]5.5092,[226]5.4868,[227]5.4691,[228]5.4543, save_imatrix: stored collected data after 228 chunks in Mistral-7B-v0.3-IMat-GGUF/imatrix.dat llama_print_timings: load time = 3151.50 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 = 180433.68 ms / 116736 tokens ( 1.55 ms per token, 646.97 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 = 183387.46 ms / 116737 tokens Final estimate: PPL = 5.4543 +/- 0.04992