|
main: build = 3024 (2b737caa) |
|
main: built with cc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0 for x86_64-linux-gnu |
|
main: seed = 1716927585 |
|
llama_model_loader: loaded meta data with 28 key-value pairs and 563 tensors from dolphin-2.9.1-mixtral-1x22b-IMat-GGUF/dolphin-2.9.1-mixtral-1x22b.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 = dolphin-2.9.1-mixtral-1x22b |
|
llama_model_loader: - kv 2: llama.block_count u32 = 56 |
|
llama_model_loader: - kv 3: llama.context_length u32 = 65536 |
|
llama_model_loader: - kv 4: llama.embedding_length u32 = 6144 |
|
llama_model_loader: - kv 5: llama.feed_forward_length u32 = 16384 |
|
llama_model_loader: - kv 6: llama.attention.head_count u32 = 48 |
|
llama_model_loader: - kv 7: llama.attention.head_count_kv u32 = 8 |
|
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: llama.expert_count u32 = 1 |
|
llama_model_loader: - kv 11: llama.expert_used_count u32 = 1 |
|
llama_model_loader: - kv 12: general.file_type u32 = 0 |
|
llama_model_loader: - kv 13: llama.vocab_size u32 = 32002 |
|
llama_model_loader: - kv 14: llama.rope.dimension_count u32 = 128 |
|
llama_model_loader: - kv 15: tokenizer.ggml.model str = llama |
|
llama_model_loader: - kv 16: tokenizer.ggml.pre str = default |
|
llama_model_loader: - kv 17: tokenizer.ggml.tokens arr[str,32002] = ["<unk>", "<s>", "</s>", "<0x00>", "<... |
|
llama_model_loader: - kv 18: tokenizer.ggml.scores arr[f32,32002] = [0.000000, 0.000000, 0.000000, 0.0000... |
|
llama_model_loader: - kv 19: tokenizer.ggml.token_type arr[i32,32002] = [2, 3, 3, 6, 6, 6, 6, 6, 6, 6, 6, 6, ... |
|
llama_model_loader: - kv 20: tokenizer.ggml.bos_token_id u32 = 1 |
|
llama_model_loader: - kv 21: tokenizer.ggml.eos_token_id u32 = 32000 |
|
llama_model_loader: - kv 22: tokenizer.ggml.unknown_token_id u32 = 0 |
|
llama_model_loader: - kv 23: tokenizer.ggml.padding_token_id u32 = 2 |
|
llama_model_loader: - kv 24: tokenizer.ggml.add_bos_token bool = true |
|
llama_model_loader: - kv 25: tokenizer.ggml.add_eos_token bool = false |
|
llama_model_loader: - kv 26: tokenizer.chat_template str = {% if not add_generation_prompt is de... |
|
llama_model_loader: - kv 27: general.quantization_version u32 = 2 |
|
llama_model_loader: - type f32: 563 tensors |
|
llm_load_vocab: special tokens definition check successful ( 261/32002 ). |
|
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 = 32002 |
|
llm_load_print_meta: n_merges = 0 |
|
llm_load_print_meta: n_ctx_train = 65536 |
|
llm_load_print_meta: n_embd = 6144 |
|
llm_load_print_meta: n_head = 48 |
|
llm_load_print_meta: n_head_kv = 8 |
|
llm_load_print_meta: n_layer = 56 |
|
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 = 6 |
|
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 = 16384 |
|
llm_load_print_meta: n_expert = 1 |
|
llm_load_print_meta: n_expert_used = 1 |
|
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_yarn_orig_ctx = 65536 |
|
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 = 22.24 B |
|
llm_load_print_meta: model size = 82.84 GiB (32.00 BPW) |
|
llm_load_print_meta: general.name = dolphin-2.9.1-mixtral-1x22b |
|
llm_load_print_meta: BOS token = 1 '<s>' |
|
llm_load_print_meta: EOS token = 32000 '<|im_end|>' |
|
llm_load_print_meta: UNK token = 0 '<unk>' |
|
llm_load_print_meta: PAD token = 2 '</s>' |
|
llm_load_print_meta: LF token = 13 '<0x0A>' |
|
llm_load_print_meta: EOT token = 32000 '<|im_end|>' |
|
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 = 0.56 MiB |
|
llm_load_tensors: offloading 15 repeating layers to GPU |
|
llm_load_tensors: offloaded 15/57 layers to GPU |
|
llm_load_tensors: CPU buffer size = 84832.05 MiB |
|
llm_load_tensors: CUDA0 buffer size = 22321.05 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: CUDA_Host KV buffer size = 82.00 MiB |
|
llama_kv_cache_init: CUDA0 KV buffer size = 30.00 MiB |
|
llama_new_context_with_model: KV self size = 112.00 MiB, K (f16): 56.00 MiB, V (f16): 56.00 MiB |
|
llama_new_context_with_model: CUDA_Host output buffer size = 0.12 MiB |
|
llama_new_context_with_model: CUDA0 compute buffer size = 848.55 MiB |
|
llama_new_context_with_model: CUDA_Host compute buffer size = 13.01 MiB |
|
llama_new_context_with_model: graph nodes = 2582 |
|
llama_new_context_with_model: graph splits = 496 |
|
|
|
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 134.661 ms |
|
compute_imatrix: computing over 228 chunks with batch_size 512 |
|
compute_imatrix: 4.52 seconds per pass - ETA 17.17 minutes |
|
[1]7.2365,[2]5.7391,[3]6.1982,[4]6.9480,[5]7.9538,[6]8.0840,[7]7.1985,[8]7.9150,[9]8.1353, |
|
save_imatrix: stored collected data after 10 chunks in dolphin-2.9.1-mixtral-1x22b-IMat-GGUF/imatrix.dat |
|
[10]8.8313,[11]8.6879,[12]8.4306,[13]8.7635,[14]7.9101,[15]8.4159,[16]8.4585,[17]8.7322,[18]8.7458,[19]8.9025, |
|
save_imatrix: stored collected data after 20 chunks in dolphin-2.9.1-mixtral-1x22b-IMat-GGUF/imatrix.dat |
|
[20]11.3244,[21]12.0510,[22]11.5303,[23]10.9382,[24]10.8757,[25]10.0453,[26]9.3982,[27]9.1256,[28]8.9954,[29]8.8928, |
|
save_imatrix: stored collected data after 30 chunks in dolphin-2.9.1-mixtral-1x22b-IMat-GGUF/imatrix.dat |
|
[30]8.9943,[31]9.1440,[32]9.3332,[33]9.3941,[34]9.5023,[35]9.1734,[36]8.9544,[37]8.7439,[38]8.6327,[39]8.5347, |
|
save_imatrix: stored collected data after 40 chunks in dolphin-2.9.1-mixtral-1x22b-IMat-GGUF/imatrix.dat |
|
[40]8.3781,[41]8.3968,[42]8.4778,[43]8.5184,[44]8.6385,[45]8.6171,[46]8.7195,[47]8.9042,[48]9.0230,[49]9.2173, |
|
save_imatrix: stored collected data after 50 chunks in dolphin-2.9.1-mixtral-1x22b-IMat-GGUF/imatrix.dat |
|
[50]9.3223,[51]9.3335,[52]9.2646,[53]9.1576,[54]8.9368,[55]9.0182,[56]9.1040,[57]9.1477,[58]9.2333,[59]9.2348, |
|
save_imatrix: stored collected data after 60 chunks in dolphin-2.9.1-mixtral-1x22b-IMat-GGUF/imatrix.dat |
|
[60]9.3561,[61]9.4115,[62]9.5052,[63]9.5562,[64]9.6042,[65]9.6625,[66]9.7168,[67]9.8172,[68]9.9023,[69]9.8814, |
|
save_imatrix: stored collected data after 70 chunks in dolphin-2.9.1-mixtral-1x22b-IMat-GGUF/imatrix.dat |
|
[70]9.8901,[71]9.8054,[72]9.7315,[73]9.6629,[74]9.5945,[75]9.5715,[76]9.5314,[77]9.4659,[78]9.3553,[79]9.3161, |
|
save_imatrix: stored collected data after 80 chunks in dolphin-2.9.1-mixtral-1x22b-IMat-GGUF/imatrix.dat |
|
[80]9.2926,[81]9.2443,[82]9.3078,[83]9.3437,[84]9.3232,[85]9.2128,[86]9.2329,[87]9.1612,[88]9.1768,[89]9.1972, |
|
save_imatrix: stored collected data after 90 chunks in dolphin-2.9.1-mixtral-1x22b-IMat-GGUF/imatrix.dat |
|
[90]9.2075,[91]9.2585,[92]9.2561,[93]9.3132,[94]9.3411,[95]9.2516,[96]9.1983,[97]9.1570,[98]9.1890,[99]9.1905, |
|
save_imatrix: stored collected data after 100 chunks in dolphin-2.9.1-mixtral-1x22b-IMat-GGUF/imatrix.dat |
|
[100]9.1427,[101]9.0664,[102]9.1635,[103]9.1604,[104]9.1381,[105]9.1065,[106]9.0728,[107]9.0552,[108]9.0485,[109]9.0668, |
|
save_imatrix: stored collected data after 110 chunks in dolphin-2.9.1-mixtral-1x22b-IMat-GGUF/imatrix.dat |
|
[110]8.9973,[111]8.9941,[112]8.9842,[113]8.9624,[114]8.9371,[115]8.9347,[116]8.9127,[117]8.8910,[118]8.8438,[119]8.8369, |
|
save_imatrix: stored collected data after 120 chunks in dolphin-2.9.1-mixtral-1x22b-IMat-GGUF/imatrix.dat |
|
[120]8.8692,[121]8.8824,[122]8.8764,[123]8.8868,[124]8.8820,[125]8.9338,[126]8.8365,[127]8.8226,[128]8.7819,[129]8.7110, |
|
save_imatrix: stored collected data after 130 chunks in dolphin-2.9.1-mixtral-1x22b-IMat-GGUF/imatrix.dat |
|
[130]8.7217,[131]8.6601,[132]8.6215,[133]8.5838,[134]8.5290,[135]8.5275,[136]8.4818,[137]8.4333,[138]8.3830,[139]8.3488, |
|
save_imatrix: stored collected data after 140 chunks in dolphin-2.9.1-mixtral-1x22b-IMat-GGUF/imatrix.dat |
|
[140]8.3645,[141]8.3625,[142]8.3288,[143]8.3221,[144]8.3070,[145]8.3191,[146]8.3197,[147]8.3032,[148]8.3223,[149]8.3213, |
|
save_imatrix: stored collected data after 150 chunks in dolphin-2.9.1-mixtral-1x22b-IMat-GGUF/imatrix.dat |
|
[150]8.3224,[151]8.3548,[152]8.3328,[153]8.3598,[154]8.4587,[155]8.5497,[156]8.6440,[157]8.6865,[158]8.7486,[159]8.8478, |
|
save_imatrix: stored collected data after 160 chunks in dolphin-2.9.1-mixtral-1x22b-IMat-GGUF/imatrix.dat |
|
[160]8.9359,[161]8.9611,[162]8.9584,[163]8.9589,[164]8.9669,[165]8.9493,[166]8.9484,[167]8.9703,[168]8.9814,[169]8.9993, |
|
save_imatrix: stored collected data after 170 chunks in dolphin-2.9.1-mixtral-1x22b-IMat-GGUF/imatrix.dat |
|
[170]8.9948,[171]9.0148,[172]9.0431,[173]9.0363,[174]9.0606,[175]9.0249,[176]9.0363,[177]9.0417,[178]9.0573,[179]9.0278, |
|
save_imatrix: stored collected data after 180 chunks in dolphin-2.9.1-mixtral-1x22b-IMat-GGUF/imatrix.dat |
|
[180]9.0490,[181]9.0296,[182]9.0082,[183]8.9940,[184]8.9806,[185]8.9842,[186]8.9818,[187]9.0149,[188]9.0114,[189]8.9626, |
|
save_imatrix: stored collected data after 190 chunks in dolphin-2.9.1-mixtral-1x22b-IMat-GGUF/imatrix.dat |
|
[190]9.0062,[191]9.0609,[192]9.1060,[193]9.1796,[194]9.2276,[195]9.2336,[196]9.2568,[197]9.2171,[198]9.2201,[199]9.2362, |
|
save_imatrix: stored collected data after 200 chunks in dolphin-2.9.1-mixtral-1x22b-IMat-GGUF/imatrix.dat |
|
[200]9.2710,[201]9.2537,[202]9.2493,[203]9.2569,[204]9.2671,[205]9.2662,[206]9.2698,[207]9.2735,[208]9.3052,[209]9.3280, |
|
save_imatrix: stored collected data after 210 chunks in dolphin-2.9.1-mixtral-1x22b-IMat-GGUF/imatrix.dat |
|
[210]9.3683,[211]9.3565,[212]9.3592,[213]9.3934,[214]9.3855,[215]9.3826,[216]9.3850,[217]9.3711,[218]9.3812,[219]9.3456, |
|
save_imatrix: stored collected data after 220 chunks in dolphin-2.9.1-mixtral-1x22b-IMat-GGUF/imatrix.dat |
|
[220]9.3661,[221]9.4195,[222]9.4675,[223]9.5259,[224]9.5537,[225]9.5597,[226]9.5231,[227]9.4933,[228]9.4802, |
|
save_imatrix: stored collected data after 228 chunks in dolphin-2.9.1-mixtral-1x22b-IMat-GGUF/imatrix.dat |
|
|
|
llama_print_timings: load time = 8055.51 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 = 1165088.42 ms / 116736 tokens ( 9.98 ms per token, 100.19 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 = 1169570.41 ms / 116737 tokens |
|
|
|
Final estimate: PPL = 9.4802 +/- 0.10941 |
|
|
