|
main: build = 2998 (9588f196) |
|
main: built with cc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0 for x86_64-linux-gnu |
|
main: seed = 1716661663 |
|
llama_model_loader: loaded meta data with 26 key-value pairs and 291 tensors from RoMistral-7b-Instruct-IMat-GGUF/RoMistral-7b-Instruct.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 = RoMistral-7b-Instruct |
|
llama_model_loader: - kv 2: llama.block_count u32 = 32 |
|
llama_model_loader: - kv 3: llama.context_length u32 = 32768 |
|
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 = 10000.000000 |
|
llama_model_loader: - kv 9: llama.attention.layer_norm_rms_epsilon f32 = 0.000010 |
|
llama_model_loader: - kv 10: general.file_type u32 = 0 |
|
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] = ["<unk>", "<s>", "</s>", "<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.padding_token_id u32 = 0 |
|
llama_model_loader: - kv 22: tokenizer.ggml.add_bos_token bool = true |
|
llama_model_loader: - kv 23: tokenizer.ggml.add_eos_token bool = false |
|
llama_model_loader: - kv 24: tokenizer.chat_template str = {% if messages[0]['role'] == 'system'... |
|
llama_model_loader: - kv 25: general.quantization_version u32 = 2 |
|
llama_model_loader: - type f32: 291 tensors |
|
llm_load_vocab: special tokens definition check successful ( 259/32000 ). |
|
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 = 32768 |
|
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 = 10000.0 |
|
llm_load_print_meta: freq_scale_train = 1 |
|
llm_load_print_meta: n_yarn_orig_ctx = 32768 |
|
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 = 7B |
|
llm_load_print_meta: model ftype = all F32 |
|
llm_load_print_meta: model params = 7.24 B |
|
llm_load_print_meta: model size = 26.98 GiB (32.00 BPW) |
|
llm_load_print_meta: general.name = RoMistral-7b-Instruct |
|
llm_load_print_meta: BOS token = 1 '<s>' |
|
llm_load_print_meta: EOS token = 2 '</s>' |
|
llm_load_print_meta: UNK token = 0 '<unk>' |
|
llm_load_print_meta: PAD token = 0 '<unk>' |
|
llm_load_print_meta: LF token = 13 '<0x0A>' |
|
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 20 repeating layers to GPU |
|
llm_load_tensors: offloaded 20/33 layers to GPU |
|
llm_load_tensors: CPU buffer size = 27625.02 MiB |
|
llm_load_tensors: CUDA0 buffer size = 16640.62 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 = 10000.0 |
|
llama_new_context_with_model: freq_scale = 1 |
|
llama_kv_cache_init: CUDA_Host KV buffer size = 24.00 MiB |
|
llama_kv_cache_init: CUDA0 KV buffer size = 40.00 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 = 0.12 MiB |
|
llama_new_context_with_model: CUDA0 compute buffer size = 570.50 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 = 136 |
|
|
|
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 133.016 ms |
|
compute_imatrix: computing over 228 chunks with batch_size 512 |
|
compute_imatrix: 1.05 seconds per pass - ETA 4.00 minutes |
|
[1]3.8877,[2]2.8849,[3]2.9218,[4]3.0824,[5]3.5000,[6]3.4194,[7]3.1548,[8]3.5868,[9]3.7105, |
|
save_imatrix: stored collected data after 10 chunks in RoMistral-7b-Instruct-IMat-GGUF/imatrix.dat |
|
[10]4.0932,[11]4.2695,[12]4.1776,[13]4.4341,[14]4.2343,[15]4.5787,[16]4.7136,[17]4.9326,[18]5.0514,[19]5.2174, |
|
save_imatrix: stored collected data after 20 chunks in RoMistral-7b-Instruct-IMat-GGUF/imatrix.dat |
|
[20]5.3147,[21]5.4322,[22]5.3098,[23]5.1083,[24]5.2085,[25]4.9603,[26]4.7743,[27]4.6601,[28]4.5907,[29]4.5878, |
|
save_imatrix: stored collected data after 30 chunks in RoMistral-7b-Instruct-IMat-GGUF/imatrix.dat |
|
[30]4.6888,[31]4.8105,[32]4.9218,[33]4.9428,[34]5.0032,[35]4.8317,[36]4.7306,[37]4.6690,[38]4.6691,[39]4.6583, |
|
save_imatrix: stored collected data after 40 chunks in RoMistral-7b-Instruct-IMat-GGUF/imatrix.dat |
|
[40]4.6157,[41]4.6577,[42]4.5956,[43]4.6585,[44]4.7544,[45]4.7648,[46]4.8477,[47]4.9766,[48]5.0826,[49]5.2213, |
|
save_imatrix: stored collected data after 50 chunks in RoMistral-7b-Instruct-IMat-GGUF/imatrix.dat |
|
[50]5.3086,[51]5.3254,[52]5.2894,[53]5.2526,[54]5.1617,[55]5.2172,[56]5.2660,[57]5.3169,[58]5.3619,[59]5.3794, |
|
save_imatrix: stored collected data after 60 chunks in RoMistral-7b-Instruct-IMat-GGUF/imatrix.dat |
|
[60]5.4490,[61]5.4817,[62]5.5301,[63]5.5509,[64]5.5718,[65]5.6034,[66]5.6377,[67]5.6796,[68]5.7246,[69]5.7424, |
|
save_imatrix: stored collected data after 70 chunks in RoMistral-7b-Instruct-IMat-GGUF/imatrix.dat |
|
[70]5.7699,[71]5.7311,[72]5.6920,[73]5.6643,[74]5.6370,[75]5.6295,[76]5.6207,[77]5.5952,[78]5.5466,[79]5.5250, |
|
save_imatrix: stored collected data after 80 chunks in RoMistral-7b-Instruct-IMat-GGUF/imatrix.dat |
|
[80]5.5204,[81]5.4951,[82]5.5372,[83]5.5733,[84]5.5915,[85]5.5357,[86]5.5526,[87]5.5173,[88]5.4581,[89]5.4367, |
|
save_imatrix: stored collected data after 90 chunks in RoMistral-7b-Instruct-IMat-GGUF/imatrix.dat |
|
[90]5.4182,[91]5.4266,[92]5.4204,[93]5.4340,[94]5.4245,[95]5.3786,[96]5.3487,[97]5.3444,[98]5.3710,[99]5.3872, |
|
save_imatrix: stored collected data after 100 chunks in RoMistral-7b-Instruct-IMat-GGUF/imatrix.dat |
|
[100]5.3820,[101]5.3588,[102]5.3380,[103]5.3497,[104]5.3471,[105]5.3324,[106]5.3172,[107]5.3193,[108]5.3282,[109]5.3468, |
|
save_imatrix: stored collected data after 110 chunks in RoMistral-7b-Instruct-IMat-GGUF/imatrix.dat |
|
[110]5.3331,[111]5.3387,[112]5.3356,[113]5.3319,[114]5.3229,[115]5.3285,[116]5.3286,[117]5.3240,[118]5.3004,[119]5.3092, |
|
save_imatrix: stored collected data after 120 chunks in RoMistral-7b-Instruct-IMat-GGUF/imatrix.dat |
|
[120]5.3323,[121]5.3414,[122]5.3397,[123]5.3500,[124]5.3555,[125]5.3770,[126]5.3317,[127]5.3285,[128]5.3092,[129]5.2853, |
|
save_imatrix: stored collected data after 130 chunks in RoMistral-7b-Instruct-IMat-GGUF/imatrix.dat |
|
[130]5.3035,[131]5.2810,[132]5.2575,[133]5.2337,[134]5.2112,[135]5.1878,[136]5.1662,[137]5.1470,[138]5.1285,[139]5.1187, |
|
save_imatrix: stored collected data after 140 chunks in RoMistral-7b-Instruct-IMat-GGUF/imatrix.dat |
|
[140]5.1144,[141]5.1067,[142]5.0870,[143]5.0821,[144]5.0707,[145]5.0621,[146]5.0523,[147]5.0415,[148]5.0361,[149]5.0243, |
|
save_imatrix: stored collected data after 150 chunks in RoMistral-7b-Instruct-IMat-GGUF/imatrix.dat |
|
[150]5.0167,[151]5.0310,[152]5.0099,[153]5.0113,[154]5.0352,[155]5.0594,[156]5.0710,[157]5.0867,[158]5.1064,[159]5.1386, |
|
save_imatrix: stored collected data after 160 chunks in RoMistral-7b-Instruct-IMat-GGUF/imatrix.dat |
|
[160]5.1558,[161]5.1721,[162]5.1779,[163]5.1876,[164]5.2073,[165]5.2074,[166]5.2147,[167]5.2312,[168]5.2416,[169]5.2557, |
|
save_imatrix: stored collected data after 170 chunks in RoMistral-7b-Instruct-IMat-GGUF/imatrix.dat |
|
[170]5.2549,[171]5.2719,[172]5.2864,[173]5.2890,[174]5.3031,[175]5.2926,[176]5.3103,[177]5.3173,[178]5.3333,[179]5.3290, |
|
save_imatrix: stored collected data after 180 chunks in RoMistral-7b-Instruct-IMat-GGUF/imatrix.dat |
|
[180]5.3447,[181]5.3459,[182]5.3448,[183]5.3432,[184]5.3403,[185]5.3501,[186]5.3569,[187]5.3761,[188]5.3778,[189]5.3583, |
|
save_imatrix: stored collected data after 190 chunks in RoMistral-7b-Instruct-IMat-GGUF/imatrix.dat |
|
[190]5.3892,[191]5.4220,[192]5.4505,[193]5.4976,[194]5.5319,[195]5.5412,[196]5.5505,[197]5.5339,[198]5.5383,[199]5.5554, |
|
save_imatrix: stored collected data after 200 chunks in RoMistral-7b-Instruct-IMat-GGUF/imatrix.dat |
|
[200]5.5805,[201]5.5785,[202]5.5773,[203]5.5839,[204]5.5960,[205]5.6002,[206]5.6075,[207]5.6168,[208]5.6234,[209]5.6366, |
|
save_imatrix: stored collected data after 210 chunks in RoMistral-7b-Instruct-IMat-GGUF/imatrix.dat |
|
[210]5.6586,[211]5.6522,[212]5.6532,[213]5.6479,[214]5.6427,[215]5.6380,[216]5.6319,[217]5.6298,[218]5.6429,[219]5.6303, |
|
save_imatrix: stored collected data after 220 chunks in RoMistral-7b-Instruct-IMat-GGUF/imatrix.dat |
|
[220]5.6409,[221]5.6726,[222]5.6918,[223]5.7166,[224]5.7343,[225]5.7378,[226]5.7150,[227]5.6980,[228]5.6837, |
|
save_imatrix: stored collected data after 228 chunks in RoMistral-7b-Instruct-IMat-GGUF/imatrix.dat |
|
|
|
llama_print_timings: load time = 2939.59 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 = 230609.75 ms / 116736 tokens ( 1.98 ms per token, 506.21 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 = 233168.01 ms / 116737 tokens |
|
|
|
Final estimate: PPL = 5.6837 +/- 0.05229 |
|
|
