llama_model_loader: loaded meta data with 24 key-value pairs and 363 tensors from llm-compiler-13b-ftd-IMat-GGUF/llm-compiler-13b-ftd.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.name str = llm-compiler-13b-ftd llama_model_loader: - kv 2: llama.block_count u32 = 40 llama_model_loader: - kv 3: llama.context_length u32 = 16384 llama_model_loader: - kv 4: llama.embedding_length u32 = 5120 llama_model_loader: - kv 5: llama.feed_forward_length u32 = 13824 llama_model_loader: - kv 6: llama.attention.head_count u32 = 40 llama_model_loader: - kv 7: llama.attention.head_count_kv u32 = 40 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 = 7 llama_model_loader: - kv 11: llama.vocab_size u32 = 32000 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,32000] = ["", "~~", "~~", "<0x00>", "<... llama_model_loader: - kv 16: tokenizer.ggml.scores arr[f32,32000] = [0.000000, 0.000000, 0.000000, 0.0000... llama_model_loader: - kv 17: tokenizer.ggml.token_type arr[i32,32000] = [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.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: 81 tensors llama_model_loader: - type q8_0: 282 tensors llm_load_vocab: special tokens cache size = 259 llm_load_vocab: token to piece cache size = 0.1684 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 = 32000 llm_load_print_meta: n_merges = 0 llm_load_print_meta: n_ctx_train = 16384 llm_load_print_meta: n_embd = 5120 llm_load_print_meta: n_head = 40 llm_load_print_meta: n_head_kv = 40 llm_load_print_meta: n_layer = 40 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 = 5120 llm_load_print_meta: n_embd_v_gqa = 5120 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 = 13824 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 = 16384 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 = 13B llm_load_print_meta: model ftype = Q8_0 llm_load_print_meta: model params = 13.02 B llm_load_print_meta: model size = 12.88 GiB (8.50 BPW) llm_load_print_meta: general.name = llm-compiler-13b-ftd 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>' 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 = 0.34 MiB llm_load_tensors: offloading 40 repeating layers to GPU llm_load_tensors: offloading non-repeating layers to GPU llm_load_tensors: offloaded 41/41 layers to GPU llm_load_tensors: CPU buffer size = 166.02 MiB llm_load_tensors: CUDA0 buffer size = 13023.85 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: CUDA0 KV buffer size = 400.00 MiB llama_new_context_with_model: KV self size = 400.00 MiB, K (f16): 200.00 MiB, V (f16): 200.00 MiB llama_new_context_with_model: CUDA_Host output buffer size = 0.12 MiB llama_new_context_with_model: CUDA0 compute buffer size = 85.00 MiB llama_new_context_with_model: CUDA_Host compute buffer size = 11.01 MiB llama_new_context_with_model: graph nodes = 1286 llama_new_context_with_model: graph splits = 2 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 94.794 ms compute_imatrix: computing over 151 chunks with batch_size 512 compute_imatrix: 0.94 seconds per pass - ETA 2.35 minutes [1]7.9287,[2]5.9126,[3]5.9611,[4]7.0636,[5]8.0817,[6]8.2143,[7]7.6573,[8]8.3965,[9]8.6270, save_imatrix: stored collected data after 10 chunks in llm-compiler-13b-ftd-IMat-GGUF/imatrix.dat [10]9.1857,[11]9.0460,[12]7.9303,[13]7.7470,[14]8.0809,[15]8.6030,[16]8.6774,[17]9.0949,[18]9.3430,[19]9.5302, save_imatrix: stored collected data after 20 chunks in llm-compiler-13b-ftd-IMat-GGUF/imatrix.dat [20]9.6224,[21]9.9279,[22]9.4080,[23]8.9325,[24]9.0198,[25]9.1096,[26]9.0443,[27]8.8646,[28]9.0256,[29]9.2171, save_imatrix: stored collected data after 30 chunks in llm-compiler-13b-ftd-IMat-GGUF/imatrix.dat [30]9.3855,[31]9.3655,[32]9.5428,[33]9.6735,[34]9.9463,[35]10.0375,[36]9.9344,[37]9.4485,[38]9.1315,[39]9.0472, save_imatrix: stored collected data after 40 chunks in llm-compiler-13b-ftd-IMat-GGUF/imatrix.dat [40]8.9374,[41]8.8264,[42]8.7242,[43]8.5519,[44]8.4773,[45]8.3585,[46]8.3491,[47]8.3841,[48]8.4458,[49]8.5460, save_imatrix: stored collected data after 50 chunks in llm-compiler-13b-ftd-IMat-GGUF/imatrix.dat [50]8.5607,[51]8.7634,[52]8.9546,[53]9.1433,[54]9.3270,[55]9.4179,[56]9.3549,[57]9.2779,[58]9.3445,[59]9.4256, save_imatrix: stored collected data after 60 chunks in llm-compiler-13b-ftd-IMat-GGUF/imatrix.dat [60]9.5378,[61]9.3954,[62]9.4207,[63]9.5214,[64]9.6259,[65]9.7009,[66]9.7442,[67]9.8208,[68]9.8889,[69]9.8691, save_imatrix: stored collected data after 70 chunks in llm-compiler-13b-ftd-IMat-GGUF/imatrix.dat [70]9.8862,[71]9.8811,[72]9.8068,[73]9.7545,[74]9.6899,[75]9.6692,[76]9.6875,[77]9.7062,[78]9.6814,[79]9.6965, save_imatrix: stored collected data after 80 chunks in llm-compiler-13b-ftd-IMat-GGUF/imatrix.dat [80]9.7250,[81]9.7054,[82]9.6918,[83]9.6268,[84]9.6397,[85]9.6431,[86]9.6338,[87]9.6605,[88]9.6553,[89]9.6398, save_imatrix: stored collected data after 90 chunks in llm-compiler-13b-ftd-IMat-GGUF/imatrix.dat [90]9.6073,[91]9.6137,[92]9.5571,[93]9.5437,[94]9.5084,[95]9.4562,[96]9.4916,[97]9.4880,[98]9.5023,[99]9.4801, save_imatrix: stored collected data after 100 chunks in llm-compiler-13b-ftd-IMat-GGUF/imatrix.dat [100]9.4738,[101]9.5066,[102]9.4704,[103]9.4313,[104]9.4120,[105]9.4428,[106]9.4445,[107]9.4561,[108]9.4832,[109]9.3903, save_imatrix: stored collected data after 110 chunks in llm-compiler-13b-ftd-IMat-GGUF/imatrix.dat [110]9.3133,[111]9.2322,[112]9.1450,[113]9.0593,[114]8.9827,[115]8.9079,[116]8.8358,[117]8.7795,[118]8.8079,[119]8.8186, save_imatrix: stored collected data after 120 chunks in llm-compiler-13b-ftd-IMat-GGUF/imatrix.dat [120]8.8760,[121]8.9375,[122]9.0009,[123]9.0650,[124]9.1673,[125]9.2741,[126]9.2882,[127]9.3165,[128]9.2212,[129]9.2221, save_imatrix: stored collected data after 130 chunks in llm-compiler-13b-ftd-IMat-GGUF/imatrix.dat [130]9.2053,[131]9.2014,[132]9.1704,[133]9.1621,[134]9.1797,[135]9.2075,[136]9.1946,[137]9.1891,[138]9.1966,[139]9.2162, save_imatrix: stored collected data after 140 chunks in llm-compiler-13b-ftd-IMat-GGUF/imatrix.dat [140]9.2363,[141]9.2379,[142]9.2320,[143]9.2133,[144]9.1826,[145]9.1988,[146]9.2285,[147]9.2656,[148]9.2995,[149]9.3494, save_imatrix: stored collected data after 150 chunks in llm-compiler-13b-ftd-IMat-GGUF/imatrix.dat [150]9.3979,[151]9.4413, save_imatrix: stored collected data after 151 chunks in llm-compiler-13b-ftd-IMat-GGUF/imatrix.dat llama_print_timings: load time = 3169.00 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 = 132055.95 ms / 77312 tokens ( 1.71 ms per token, 585.45 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 = 134723.77 ms / 77313 tokens Final estimate: PPL = 9.4413 +/- 0.12890