File size: 10,243 Bytes
b029265 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 |
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
|