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 = 1716766865 llama_model_loader: loaded meta data with 22 key-value pairs and 291 tensors from Llama3-ChatQA-1.5-8B-IMat-GGUF/Llama3-ChatQA-1.5-8B.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 = Llama3-ChatQA-1.5-8B llama_model_loader: - kv 2: llama.block_count u32 = 32 llama_model_loader: - kv 3: llama.context_length u32 = 8192 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 = 500000.000000 llama_model_loader: - kv 9: llama.attention.layer_norm_rms_epsilon f32 = 0.000010 llama_model_loader: - kv 10: general.file_type u32 = 1 llama_model_loader: - kv 11: llama.vocab_size u32 = 128256 llama_model_loader: - kv 12: llama.rope.dimension_count u32 = 128 llama_model_loader: - kv 13: tokenizer.ggml.model str = gpt2 llama_model_loader: - kv 14: tokenizer.ggml.pre str = smaug-bpe llama_model_loader: - kv 15: tokenizer.ggml.tokens arr[str,128256] = ["!", "\"", "#", "$", "%", "&", "'", ... llama_model_loader: - kv 16: tokenizer.ggml.token_type arr[i32,128256] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... llama_model_loader: - kv 17: tokenizer.ggml.merges arr[str,280147] = ["Ġ Ġ", "Ġ ĠĠĠ", "ĠĠ ĠĠ", "... llama_model_loader: - kv 18: tokenizer.ggml.bos_token_id u32 = 128000 llama_model_loader: - kv 19: tokenizer.ggml.eos_token_id u32 = 128001 llama_model_loader: - kv 20: tokenizer.chat_template str = {{ bos_token }}{%- if messages[0]['ro... llama_model_loader: - kv 21: general.quantization_version u32 = 2 llama_model_loader: - type f32: 65 tensors llama_model_loader: - type f16: 226 tensors llm_load_vocab: special tokens definition check successful ( 256/128256 ). llm_load_print_meta: format = GGUF V3 (latest) llm_load_print_meta: arch = llama llm_load_print_meta: vocab type = BPE llm_load_print_meta: n_vocab = 128256 llm_load_print_meta: n_merges = 280147 llm_load_print_meta: n_ctx_train = 8192 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 = 500000.0 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 = 8B llm_load_print_meta: model ftype = F16 llm_load_print_meta: model params = 8.03 B llm_load_print_meta: model size = 14.96 GiB (16.00 BPW) llm_load_print_meta: general.name = Llama3-ChatQA-1.5-8B llm_load_print_meta: BOS token = 128000 '<|begin_of_text|>' llm_load_print_meta: EOS token = 128001 '<|end_of_text|>' llm_load_print_meta: LF token = 128 'Ä' llm_load_print_meta: EOT token = 128009 '<|eot_id|>' 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 32 repeating layers to GPU llm_load_tensors: offloading non-repeating layers to GPU llm_load_tensors: offloaded 33/33 layers to GPU llm_load_tensors: CPU buffer size = 1002.00 MiB llm_load_tensors: CUDA0 buffer size = 14315.02 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 = 500000.0 llama_new_context_with_model: freq_scale = 1 llama_kv_cache_init: CUDA0 KV buffer size = 64.00 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 = 0.49 MiB llama_new_context_with_model: CUDA0 compute buffer size = 258.50 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 = 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 198.763 ms compute_imatrix: computing over 189 chunks with batch_size 512 compute_imatrix: 0.55 seconds per pass - ETA 1.73 minutes [1]5.5379,[2]4.2604,[3]3.8769,[4]4.7854,[5]4.8156,[6]4.0973,[7]4.4325,[8]4.8732,[9]5.0451, save_imatrix: stored collected data after 10 chunks in Llama3-ChatQA-1.5-8B-IMat-GGUF/imatrix.dat [10]5.0631,[11]5.5003,[12]5.3789,[13]5.8285,[14]6.2268,[15]6.4722,[16]6.8534,[17]7.2681,[18]7.4198,[19]7.0782, save_imatrix: stored collected data after 20 chunks in Llama3-ChatQA-1.5-8B-IMat-GGUF/imatrix.dat [20]6.9814,[21]6.8167,[22]6.4535,[23]6.2212,[24]6.0214,[25]6.2230,[26]6.3250,[27]6.4800,[28]6.4423,[29]6.1705, save_imatrix: stored collected data after 30 chunks in Llama3-ChatQA-1.5-8B-IMat-GGUF/imatrix.dat [30]5.9908,[31]5.9112,[32]5.8855,[33]5.8623,[34]5.8742,[35]5.9606,[36]6.0615,[37]6.1978,[38]6.2610,[39]6.3864, save_imatrix: stored collected data after 40 chunks in Llama3-ChatQA-1.5-8B-IMat-GGUF/imatrix.dat [40]6.5592,[41]6.7632,[42]6.8866,[43]7.0488,[44]7.0439,[45]7.0809,[46]7.1623,[47]7.2816,[48]7.3168,[49]7.3888, save_imatrix: stored collected data after 50 chunks in Llama3-ChatQA-1.5-8B-IMat-GGUF/imatrix.dat [50]7.4099,[51]7.3661,[52]7.1987,[53]7.1230,[54]7.1023,[55]6.9628,[56]6.8335,[57]6.8427,[58]6.9167,[59]7.0103, save_imatrix: stored collected data after 60 chunks in Llama3-ChatQA-1.5-8B-IMat-GGUF/imatrix.dat [60]7.0712,[61]7.0300,[62]6.9374,[63]6.8442,[64]6.7532,[65]6.6802,[66]6.5717,[67]6.4531,[68]6.4243,[69]6.3643, save_imatrix: stored collected data after 70 chunks in Llama3-ChatQA-1.5-8B-IMat-GGUF/imatrix.dat [70]6.3744,[71]6.4125,[72]6.4348,[73]6.4352,[74]6.4698,[75]6.4065,[76]6.2850,[77]6.1718,[78]6.0963,[79]5.9844, save_imatrix: stored collected data after 80 chunks in Llama3-ChatQA-1.5-8B-IMat-GGUF/imatrix.dat [80]5.8879,[81]5.7904,[82]5.7283,[83]5.6832,[84]5.7078,[85]5.7533,[86]5.7654,[87]5.7501,[88]5.7441,[89]5.7588, save_imatrix: stored collected data after 90 chunks in Llama3-ChatQA-1.5-8B-IMat-GGUF/imatrix.dat [90]5.7898,[91]5.7871,[92]5.7989,[93]5.8200,[94]5.8462,[95]5.8372,[96]5.8650,[97]5.8730,[98]5.8785,[99]5.8935, save_imatrix: stored collected data after 100 chunks in Llama3-ChatQA-1.5-8B-IMat-GGUF/imatrix.dat [100]5.8926,[101]5.8860,[102]5.8938,[103]5.9202,[104]5.9412,[105]5.9380,[106]5.9664,[107]5.9930,[108]5.9533,[109]5.9600, save_imatrix: stored collected data after 110 chunks in Llama3-ChatQA-1.5-8B-IMat-GGUF/imatrix.dat [110]5.9524,[111]5.9278,[112]5.9181,[113]5.8898,[114]5.8573,[115]5.8296,[116]5.8003,[117]5.7703,[118]5.7423,[119]5.7852, save_imatrix: stored collected data after 120 chunks in Llama3-ChatQA-1.5-8B-IMat-GGUF/imatrix.dat [120]5.8020,[121]5.8241,[122]5.8670,[123]5.8976,[124]5.9497,[125]6.0091,[126]6.0623,[127]6.1093,[128]6.1734,[129]6.2466, save_imatrix: stored collected data after 130 chunks in Llama3-ChatQA-1.5-8B-IMat-GGUF/imatrix.dat [130]6.2292,[131]6.2497,[132]6.2611,[133]6.2829,[134]6.2720,[135]6.2806,[136]6.3134,[137]6.3258,[138]6.3446,[139]6.3687, save_imatrix: stored collected data after 140 chunks in Llama3-ChatQA-1.5-8B-IMat-GGUF/imatrix.dat [140]6.3830,[141]6.3885,[142]6.4090,[143]6.3848,[144]6.4083,[145]6.4348,[146]6.4522,[147]6.4599,[148]6.4739,[149]6.4918, save_imatrix: stored collected data after 150 chunks in Llama3-ChatQA-1.5-8B-IMat-GGUF/imatrix.dat [150]6.4795,[151]6.4745,[152]6.4854,[153]6.4931,[154]6.5394,[155]6.5287,[156]6.5323,[157]6.5733,[158]6.6175,[159]6.6809, save_imatrix: stored collected data after 160 chunks in Llama3-ChatQA-1.5-8B-IMat-GGUF/imatrix.dat [160]6.7411,[161]6.7571,[162]6.7753,[163]6.7901,[164]6.7885,[165]6.8196,[166]6.8253,[167]6.8271,[168]6.8374,[169]6.8606, save_imatrix: stored collected data after 170 chunks in Llama3-ChatQA-1.5-8B-IMat-GGUF/imatrix.dat [170]6.8616,[171]6.8588,[172]6.8711,[173]6.8457,[174]6.8452,[175]6.8364,[176]6.8381,[177]6.8440,[178]6.8483,[179]6.8433, save_imatrix: stored collected data after 180 chunks in Llama3-ChatQA-1.5-8B-IMat-GGUF/imatrix.dat [180]6.8314,[181]6.8437,[182]6.8311,[183]6.8107,[184]6.7766,[185]6.7852,[186]6.7738,[187]6.7698,[188]6.7375,[189]6.7122, save_imatrix: stored collected data after 189 chunks in Llama3-ChatQA-1.5-8B-IMat-GGUF/imatrix.dat llama_print_timings: load time = 2209.94 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 = 77902.33 ms / 96768 tokens ( 0.81 ms per token, 1242.17 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 = 80773.33 ms / 96769 tokens Final estimate: PPL = 6.7122 +/- 0.07586