main: build = 3051 (5921b8f0) main: built with cc (Ubuntu 13.2.0-23ubuntu4) 13.2.0 for x86_64-linux-gnu main: seed = 1717156302 llama_model_loader: loaded meta data with 24 key-value pairs and 255 tensors from neo_7b-IMat-GGUF/neo_7b.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 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 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 21 repeating layers to GPU llm_load_tensors: offloaded 21/29 layers to GPU llm_load_tensors: CPU buffer size = 29730.67 MiB llm_load_tensors: CUDA0 buffer size = 21168.49 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 = 42.00 MiB llama_kv_cache_init: CUDA0 KV buffer size = 126.00 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 = 81 system_info: n_threads = 25 / 32 | AVX = 1 | AVX_VNNI = 0 | AVX2 = 1 | AVX512 = 0 | AVX512_VBMI = 0 | AVX512_VNNI = 0 | AVX512_BF16 = 0 | 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 107.299 ms compute_imatrix: computing over 149 chunks with batch_size 512 compute_imatrix: 1.34 seconds per pass - ETA 3.32 minutes [1]5.5510,[2]4.2655,[3]4.5181,[4]4.7129,[5]4.6112,[6]4.4538,[7]3.9156,[8]3.9898,[9]4.0393, save_imatrix: stored collected data after 10 chunks in neo_7b-IMat-GGUF/imatrix.dat [10]4.3880,[11]4.3659,[12]4.1373,[13]4.4218,[14]4.8427,[15]5.0896,[16]5.4727,[17]5.7076,[18]5.9118,[19]6.0551, save_imatrix: stored collected data after 20 chunks in neo_7b-IMat-GGUF/imatrix.dat [20]6.3265,[21]6.1195,[22]6.1174,[23]6.1416,[24]6.2290,[25]6.1717,[26]6.1541,[27]6.2767,[28]6.4483,[29]6.6253, save_imatrix: stored collected data after 30 chunks in neo_7b-IMat-GGUF/imatrix.dat [30]6.6419,[31]6.7306,[32]6.8220,[33]6.8508,[34]6.7559,[35]6.6592,[36]6.3890,[37]6.2370,[38]5.9876,[39]5.9491, save_imatrix: stored collected data after 40 chunks in neo_7b-IMat-GGUF/imatrix.dat [40]5.9009,[41]5.8543,[42]5.8175,[43]5.8264,[44]5.7960,[45]5.7330,[46]5.7837,[47]5.8467,[48]5.9655,[49]5.9876, save_imatrix: stored collected data after 50 chunks in neo_7b-IMat-GGUF/imatrix.dat [50]6.1532,[51]6.3050,[52]6.4750,[53]6.6240,[54]6.7573,[55]6.6985,[56]6.6797,[57]6.7828,[58]6.8664,[59]6.8508, save_imatrix: stored collected data after 60 chunks in neo_7b-IMat-GGUF/imatrix.dat [60]6.8056,[61]6.8193,[62]6.8890,[63]6.9615,[64]7.0462,[65]7.0945,[66]7.1400,[67]7.1548,[68]7.1813,[69]7.1861, save_imatrix: stored collected data after 70 chunks in neo_7b-IMat-GGUF/imatrix.dat [70]7.1570,[71]7.0742,[72]6.9900,[73]6.9825,[74]7.0361,[75]7.0688,[76]7.0267,[77]7.0049,[78]7.0124,[79]6.9808, save_imatrix: stored collected data after 80 chunks in neo_7b-IMat-GGUF/imatrix.dat [80]6.9087,[81]6.7737,[82]6.6562,[83]6.6602,[84]6.6616,[85]6.6651,[86]6.6312,[87]6.6456,[88]6.6331,[89]6.6257, save_imatrix: stored collected data after 90 chunks in neo_7b-IMat-GGUF/imatrix.dat [90]6.6105,[91]6.5855,[92]6.5724,[93]6.5476,[94]6.4786,[95]6.5129,[96]6.5639,[97]6.5803,[98]6.5558,[99]6.5736, save_imatrix: stored collected data after 100 chunks in neo_7b-IMat-GGUF/imatrix.dat [100]6.6030,[101]6.5436,[102]6.4977,[103]6.4963,[104]6.5190,[105]6.5218,[106]6.5436,[107]6.5605,[108]6.5171,[109]6.4749, save_imatrix: stored collected data after 110 chunks in neo_7b-IMat-GGUF/imatrix.dat [110]6.4291,[111]6.3778,[112]6.3291,[113]6.2819,[114]6.2431,[115]6.2022,[116]6.1790,[117]6.1910,[118]6.2014,[119]6.2657, save_imatrix: stored collected data after 120 chunks in neo_7b-IMat-GGUF/imatrix.dat [120]6.3278,[121]6.3850,[122]6.4316,[123]6.5052,[124]6.4924,[125]6.5206,[126]6.5060,[127]6.5299,[128]6.5179,[129]6.5090, save_imatrix: stored collected data after 130 chunks in neo_7b-IMat-GGUF/imatrix.dat [130]6.4801,[131]6.4507,[132]6.4851,[133]6.5358,[134]6.5337,[135]6.5396,[136]6.5504,[137]6.5770,[138]6.5847,[139]6.5964, save_imatrix: stored collected data after 140 chunks in neo_7b-IMat-GGUF/imatrix.dat [140]6.6219,[141]6.6355,[142]6.6113,[143]6.6594,[144]6.6952,[145]6.7260,[146]6.7622,[147]6.7856,[148]6.8268,[149]6.8470, save_imatrix: stored collected data after 149 chunks in neo_7b-IMat-GGUF/imatrix.dat llama_print_timings: load time = 4232.51 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 = 173273.88 ms / 76288 tokens ( 2.27 ms per token, 440.27 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 = 177120.13 ms / 76289 tokens Final estimate: PPL = 6.8470 +/- 0.08549