llama_model_loader: loaded meta data with 23 key-value pairs and 283 tensors from codegeex4-all-9b-IMat-GGUF/codegeex4-all-9b.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 = chatglm llama_model_loader: - kv 1: general.name str = codegeex4-all-9b llama_model_loader: - kv 2: chatglm.context_length u32 = 131072 llama_model_loader: - kv 3: chatglm.embedding_length u32 = 4096 llama_model_loader: - kv 4: chatglm.feed_forward_length u32 = 13696 llama_model_loader: - kv 5: chatglm.block_count u32 = 40 llama_model_loader: - kv 6: chatglm.attention.head_count u32 = 32 llama_model_loader: - kv 7: chatglm.attention.head_count_kv u32 = 2 llama_model_loader: - kv 8: chatglm.attention.layer_norm_rms_epsilon f32 = 0.000010 llama_model_loader: - kv 9: general.file_type u32 = 7 llama_model_loader: - kv 10: chatglm.rope.dimension_count u32 = 64 llama_model_loader: - kv 11: tokenizer.ggml.add_bos_token bool = false llama_model_loader: - kv 12: chatglm.rope.freq_base f32 = 500.000000 llama_model_loader: - kv 13: tokenizer.ggml.model str = gpt2 llama_model_loader: - kv 14: tokenizer.ggml.pre str = chatglm-bpe llama_model_loader: - kv 15: tokenizer.ggml.tokens arr[str,151552] = ["!", "\"", "#", "$", "%", "&", "'", ... llama_model_loader: - kv 16: tokenizer.ggml.token_type arr[i32,151552] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... llama_model_loader: - kv 17: tokenizer.ggml.merges arr[str,151073] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",... llama_model_loader: - kv 18: tokenizer.ggml.padding_token_id u32 = 151329 llama_model_loader: - kv 19: tokenizer.ggml.eos_token_id u32 = 151329 llama_model_loader: - kv 20: tokenizer.ggml.eot_token_id u32 = 151336 llama_model_loader: - kv 21: tokenizer.ggml.unknown_token_id u32 = 151329 llama_model_loader: - kv 22: general.quantization_version u32 = 2 llama_model_loader: - type f32: 121 tensors llama_model_loader: - type q8_0: 162 tensors llm_load_vocab: special tokens cache size = 223 llm_load_vocab: token to piece cache size = 0.9732 MB llm_load_print_meta: format = GGUF V3 (latest) llm_load_print_meta: arch = chatglm llm_load_print_meta: vocab type = BPE llm_load_print_meta: n_vocab = 151552 llm_load_print_meta: n_merges = 151073 llm_load_print_meta: n_ctx_train = 131072 llm_load_print_meta: n_embd = 4096 llm_load_print_meta: n_head = 32 llm_load_print_meta: n_head_kv = 2 llm_load_print_meta: n_layer = 40 llm_load_print_meta: n_rot = 64 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 = 16 llm_load_print_meta: n_embd_k_gqa = 256 llm_load_print_meta: n_embd_v_gqa = 256 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 = 13696 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 = 500.0 llm_load_print_meta: freq_scale_train = 1 llm_load_print_meta: n_ctx_orig_yarn = 131072 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 = 9B llm_load_print_meta: model ftype = Q8_0 llm_load_print_meta: model params = 9.40 B llm_load_print_meta: model size = 9.30 GiB (8.50 BPW) llm_load_print_meta: general.name = codegeex4-all-9b llm_load_print_meta: EOS token = 151329 '<|endoftext|>' llm_load_print_meta: UNK token = 151329 '<|endoftext|>' llm_load_print_meta: PAD token = 151329 '<|endoftext|>' llm_load_print_meta: LF token = 128 'Ä' llm_load_print_meta: EOT token = 151336 '<|user|>' llm_load_print_meta: max token length = 1024 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 = 0.28 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 = 629.00 MiB llm_load_tensors: CUDA0 buffer size = 8897.23 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 = 500.0 llama_new_context_with_model: freq_scale = 1 llama_kv_cache_init: CUDA0 KV buffer size = 20.00 MiB llama_new_context_with_model: KV self size = 20.00 MiB, K (f16): 10.00 MiB, V (f16): 10.00 MiB llama_new_context_with_model: CUDA_Host output buffer size = 0.58 MiB llama_new_context_with_model: CUDA0 compute buffer size = 304.00 MiB llama_new_context_with_model: CUDA_Host compute buffer size = 9.01 MiB llama_new_context_with_model: graph nodes = 1606 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 122.92 ms compute_imatrix: computing over 125 chunks with batch_size 512 compute_imatrix: 0.63 seconds per pass - ETA 1.30 minutes [1]10.3562,[2]8.4325,[3]7.7173,[4]9.6163,[5]9.6539,[6]7.8712,[7]8.6433,[8]9.1616,[9]9.5800, save_imatrix: stored collected data after 10 chunks in codegeex4-all-9b-IMat-GGUF/imatrix.dat [10]8.1854,[11]9.0245,[12]10.1385,[13]10.8339,[14]11.3143,[15]12.0013,[16]12.5792,[17]12.9746,[18]12.4173,[19]11.6037, save_imatrix: stored collected data after 20 chunks in codegeex4-all-9b-IMat-GGUF/imatrix.dat [20]11.7752,[21]12.4757,[22]12.1252,[23]12.7541,[24]12.6786,[25]13.2785,[26]13.4211,[27]13.7764,[28]14.4623,[29]14.7203, save_imatrix: stored collected data after 30 chunks in codegeex4-all-9b-IMat-GGUF/imatrix.dat [30]14.6740,[31]13.7066,[32]13.2003,[33]12.9453,[34]12.7611,[35]12.8194,[36]13.1588,[37]13.1278,[38]13.2979,[39]13.6020, save_imatrix: stored collected data after 40 chunks in codegeex4-all-9b-IMat-GGUF/imatrix.dat [40]13.6874,[41]14.4190,[42]15.0294,[43]15.5386,[44]15.8908,[45]16.1925,[46]15.9531,[47]16.1393,[48]16.2833,[49]16.2760, save_imatrix: stored collected data after 50 chunks in codegeex4-all-9b-IMat-GGUF/imatrix.dat [50]16.2608,[51]16.2804,[52]16.5852,[53]16.7566,[54]17.0592,[55]17.1281,[56]17.0877,[57]17.1132,[58]17.0390,[59]16.8126, save_imatrix: stored collected data after 60 chunks in codegeex4-all-9b-IMat-GGUF/imatrix.dat [60]16.6386,[61]16.4616,[62]16.4449,[63]16.5449,[64]16.3393,[65]16.2853,[66]16.2069,[67]16.0896,[68]16.0264,[69]15.9704, save_imatrix: stored collected data after 70 chunks in codegeex4-all-9b-IMat-GGUF/imatrix.dat [70]15.9065,[71]15.8381,[72]15.8003,[73]15.7025,[74]15.5344,[75]15.4538,[76]15.4497,[77]15.4111,[78]15.3426,[79]15.3654, save_imatrix: stored collected data after 80 chunks in codegeex4-all-9b-IMat-GGUF/imatrix.dat [80]15.3707,[81]15.2761,[82]15.3316,[83]15.3867,[84]15.0731,[85]15.0980,[86]15.0961,[87]15.1150,[88]15.1433,[89]15.1545, save_imatrix: stored collected data after 90 chunks in codegeex4-all-9b-IMat-GGUF/imatrix.dat [90]14.9914,[91]14.8445,[92]14.6858,[93]14.5355,[94]14.3787,[95]14.2427,[96]14.1677,[97]14.1794,[98]14.2049,[99]14.4170, save_imatrix: stored collected data after 100 chunks in codegeex4-all-9b-IMat-GGUF/imatrix.dat [100]14.6055,[101]14.7210,[102]14.9603,[103]15.0024,[104]15.0448,[105]14.8945,[106]14.8769,[107]14.8630,[108]14.7411,[109]14.6102, save_imatrix: stored collected data after 110 chunks in codegeex4-all-9b-IMat-GGUF/imatrix.dat [110]14.6423,[111]14.7466,[112]14.7437,[113]14.7807,[114]14.8204,[115]14.9192,[116]14.9062,[117]14.8987,[118]14.8812,[119]14.7589, save_imatrix: stored collected data after 120 chunks in codegeex4-all-9b-IMat-GGUF/imatrix.dat [120]14.9470,[121]15.1001,[122]15.2480,[123]15.3920,[124]15.5287,[125]15.6589, save_imatrix: stored collected data after 125 chunks in codegeex4-all-9b-IMat-GGUF/imatrix.dat llama_print_timings: load time = 16954.85 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 = 66214.21 ms / 64000 tokens ( 1.03 ms per token, 966.56 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 = 83433.10 ms / 64001 tokens Final estimate: PPL = 15.6589 +/- 0.26003