|
main: build = 3008 (1d8fca72) |
|
main: built with cc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0 for x86_64-linux-gnu |
|
main: seed = 1716821304 |
|
llama_model_loader: loaded meta data with 22 key-value pairs and 435 tensors from internlm2-math-plus-20b-IMat-GGUF/internlm2-math-plus-20b.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 = internlm2 |
|
llama_model_loader: - kv 1: general.name str = InternLM2 |
|
llama_model_loader: - kv 2: internlm2.context_length u32 = 8192 |
|
llama_model_loader: - kv 3: internlm2.block_count u32 = 48 |
|
llama_model_loader: - kv 4: internlm2.embedding_length u32 = 6144 |
|
llama_model_loader: - kv 5: internlm2.feed_forward_length u32 = 16384 |
|
llama_model_loader: - kv 6: internlm2.rope.freq_base f32 = 1000000.000000 |
|
llama_model_loader: - kv 7: internlm2.attention.head_count u32 = 48 |
|
llama_model_loader: - kv 8: internlm2.attention.layer_norm_rms_epsilon f32 = 0.000010 |
|
llama_model_loader: - kv 9: internlm2.attention.head_count_kv u32 = 8 |
|
llama_model_loader: - kv 10: general.file_type u32 = 0 |
|
llama_model_loader: - kv 11: tokenizer.ggml.model str = llama |
|
llama_model_loader: - kv 12: tokenizer.ggml.pre str = default |
|
llama_model_loader: - kv 13: tokenizer.ggml.tokens arr[str,92544] = ["<unk>", "<s>", "</s>", "<0x00>", "<... |
|
llama_model_loader: - kv 14: tokenizer.ggml.scores arr[f32,92544] = [0.000000, 0.000000, 0.000000, 0.0000... |
|
llama_model_loader: - kv 15: tokenizer.ggml.token_type arr[i32,92544] = [2, 3, 3, 6, 6, 6, 6, 6, 6, 6, 6, 6, ... |
|
llama_model_loader: - kv 16: tokenizer.ggml.add_space_prefix bool = false |
|
llama_model_loader: - kv 17: tokenizer.ggml.bos_token_id u32 = 1 |
|
llama_model_loader: - kv 18: tokenizer.ggml.eos_token_id u32 = 2 |
|
llama_model_loader: - kv 19: tokenizer.ggml.padding_token_id u32 = 2 |
|
llama_model_loader: - kv 20: tokenizer.chat_template str = {{ bos_token }}{% for message in mess... |
|
llama_model_loader: - kv 21: general.quantization_version u32 = 2 |
|
llama_model_loader: - type f32: 435 tensors |
|
llm_load_vocab: mismatch in special tokens definition ( 405/92544 vs 259/92544 ). |
|
llm_load_print_meta: format = GGUF V3 (latest) |
|
llm_load_print_meta: arch = internlm2 |
|
llm_load_print_meta: vocab type = SPM |
|
llm_load_print_meta: n_vocab = 92544 |
|
llm_load_print_meta: n_merges = 0 |
|
llm_load_print_meta: n_ctx_train = 8192 |
|
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 = 48 |
|
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 = 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 = 1000000.0 |
|
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 = 20B |
|
llm_load_print_meta: model ftype = all F32 |
|
llm_load_print_meta: model params = 19.86 B |
|
llm_load_print_meta: model size = 73.99 GiB (32.00 BPW) |
|
llm_load_print_meta: general.name = InternLM2 |
|
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 = 2 '</s>' |
|
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 = 0.44 MiB |
|
llm_load_tensors: offloading 13 repeating layers to GPU |
|
llm_load_tensors: offloaded 13/49 layers to GPU |
|
llm_load_tensors: CPU buffer size = 75764.27 MiB |
|
llm_load_tensors: CUDA0 buffer size = 19344.61 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 = 70.00 MiB |
|
llama_kv_cache_init: CUDA0 KV buffer size = 26.00 MiB |
|
llama_new_context_with_model: KV self size = 96.00 MiB, K (f16): 48.00 MiB, V (f16): 48.00 MiB |
|
llama_new_context_with_model: CUDA_Host output buffer size = 0.35 MiB |
|
llama_new_context_with_model: CUDA0 compute buffer size = 2361.75 MiB |
|
llama_new_context_with_model: CUDA_Host compute buffer size = 13.01 MiB |
|
llama_new_context_with_model: graph nodes = 1542 |
|
llama_new_context_with_model: graph splits = 389 |
|
|
|
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 142.103 ms |
|
compute_imatrix: computing over 209 chunks with batch_size 512 |
|
compute_imatrix: 4.03 seconds per pass - ETA 14.05 minutes |
|
[1]6.4168,[2]4.7583,[3]4.4295,[4]5.0986,[5]5.1323,[6]4.6396,[7]5.0738,[8]5.0125,[9]5.5265, |
|
save_imatrix: stored collected data after 10 chunks in internlm2-math-plus-20b-IMat-GGUF/imatrix.dat |
|
[10]5.7465,[11]6.2022,[12]6.3583,[13]7.3490,[14]7.7210,[15]8.3283,[16]8.6313,[17]9.0740,[18]8.6086,[19]8.9529, |
|
save_imatrix: stored collected data after 20 chunks in internlm2-math-plus-20b-IMat-GGUF/imatrix.dat |
|
[20]8.9419,[21]8.5055,[22]8.6022,[23]8.0311,[24]7.7363,[25]7.3791,[26]7.2925,[27]7.6497,[28]7.6530,[29]7.9480, |
|
save_imatrix: stored collected data after 30 chunks in internlm2-math-plus-20b-IMat-GGUF/imatrix.dat |
|
[30]8.1636,[31]8.1232,[32]7.7440,[33]7.5306,[34]7.4294,[35]7.3570,[36]7.2975,[37]7.4934,[38]7.6827,[39]7.8636, |
|
save_imatrix: stored collected data after 40 chunks in internlm2-math-plus-20b-IMat-GGUF/imatrix.dat |
|
[40]8.0726,[41]8.2202,[42]8.4753,[43]8.7291,[44]8.9744,[45]9.0965,[46]9.1642,[47]9.0982,[48]9.0409,[49]9.1886, |
|
save_imatrix: stored collected data after 50 chunks in internlm2-math-plus-20b-IMat-GGUF/imatrix.dat |
|
[50]9.3875,[51]9.4652,[52]9.6524,[53]9.6800,[54]9.6847,[55]9.6258,[56]9.6068,[57]9.5647,[58]9.5839,[59]9.5052, |
|
save_imatrix: stored collected data after 60 chunks in internlm2-math-plus-20b-IMat-GGUF/imatrix.dat |
|
[60]9.5133,[61]9.6788,[62]9.8504,[63]10.0499,[64]10.0346,[65]9.8966,[66]9.6476,[67]9.4465,[68]9.2346,[69]8.9941, |
|
save_imatrix: stored collected data after 70 chunks in internlm2-math-plus-20b-IMat-GGUF/imatrix.dat |
|
[70]8.8648,[71]8.6641,[72]8.4585,[73]8.2665,[74]8.3704,[75]8.4552,[76]8.4843,[77]8.4832,[78]8.5569,[79]8.5205, |
|
save_imatrix: stored collected data after 80 chunks in internlm2-math-plus-20b-IMat-GGUF/imatrix.dat |
|
[80]8.4785,[81]8.3997,[82]8.3609,[83]8.3337,[84]8.2886,[85]8.2690,[86]8.2590,[87]8.2710,[88]8.2770,[89]8.3386, |
|
save_imatrix: stored collected data after 90 chunks in internlm2-math-plus-20b-IMat-GGUF/imatrix.dat |
|
[90]8.3979,[91]8.4345,[92]8.4115,[93]8.3837,[94]8.3554,[95]8.3623,[96]8.3118,[97]8.3259,[98]8.3270,[99]8.3270, |
|
save_imatrix: stored collected data after 100 chunks in internlm2-math-plus-20b-IMat-GGUF/imatrix.dat |
|
[100]8.2891,[101]8.2983,[102]8.2687,[103]8.2242,[104]8.1843,[105]8.1753,[106]8.1447,[107]8.1064,[108]8.0839,[109]8.0914, |
|
save_imatrix: stored collected data after 110 chunks in internlm2-math-plus-20b-IMat-GGUF/imatrix.dat |
|
[110]8.1122,[111]8.0960,[112]8.1216,[113]8.1442,[114]8.1168,[115]8.0901,[116]8.1291,[117]8.1107,[118]8.1351,[119]8.0801, |
|
save_imatrix: stored collected data after 120 chunks in internlm2-math-plus-20b-IMat-GGUF/imatrix.dat |
|
[120]8.0299,[121]7.9685,[122]7.9067,[123]7.8508,[124]7.7999,[125]7.7499,[126]7.7376,[127]7.7151,[128]7.6827,[129]7.6379, |
|
save_imatrix: stored collected data after 130 chunks in internlm2-math-plus-20b-IMat-GGUF/imatrix.dat |
|
[130]7.6230,[131]7.5946,[132]7.5657,[133]7.5499,[134]7.5140,[135]7.4877,[136]7.4753,[137]7.4455,[138]7.4158,[139]7.3971, |
|
save_imatrix: stored collected data after 140 chunks in internlm2-math-plus-20b-IMat-GGUF/imatrix.dat |
|
[140]7.3603,[141]7.3353,[142]7.4085,[143]7.5300,[144]7.6907,[145]7.8070,[146]7.8090,[147]7.8345,[148]7.8578,[149]7.8940, |
|
save_imatrix: stored collected data after 150 chunks in internlm2-math-plus-20b-IMat-GGUF/imatrix.dat |
|
[150]7.9156,[151]7.9165,[152]7.9163,[153]7.9517,[154]7.9755,[155]8.0222,[156]8.0379,[157]8.0682,[158]8.1081,[159]8.1213, |
|
save_imatrix: stored collected data after 160 chunks in internlm2-math-plus-20b-IMat-GGUF/imatrix.dat |
|
[160]8.1592,[161]8.1699,[162]8.2061,[163]8.2408,[164]8.2731,[165]8.2912,[166]8.3150,[167]8.3475,[168]8.3554,[169]8.3680, |
|
save_imatrix: stored collected data after 170 chunks in internlm2-math-plus-20b-IMat-GGUF/imatrix.dat |
|
[170]8.3878,[171]8.4014,[172]8.4365,[173]8.4628,[174]8.4457,[175]8.5097,[176]8.5780,[177]8.6452,[178]8.7343,[179]8.7916, |
|
save_imatrix: stored collected data after 180 chunks in internlm2-math-plus-20b-IMat-GGUF/imatrix.dat |
|
[180]8.8209,[181]8.8049,[182]8.8278,[183]8.8614,[184]8.9148,[185]8.9235,[186]8.9347,[187]8.9465,[188]8.9759,[189]8.9848, |
|
save_imatrix: stored collected data after 190 chunks in internlm2-math-plus-20b-IMat-GGUF/imatrix.dat |
|
[190]8.9923,[191]9.0185,[192]9.0381,[193]9.0588,[194]9.0475,[195]9.0663,[196]9.0615,[197]9.0806,[198]9.0922,[199]9.1492, |
|
save_imatrix: stored collected data after 200 chunks in internlm2-math-plus-20b-IMat-GGUF/imatrix.dat |
|
[200]9.1036,[201]9.1208,[202]9.1109,[203]9.1864,[204]9.2484,[205]9.3149,[206]9.3636,[207]9.4050,[208]9.3653,[209]9.3305, |
|
save_imatrix: stored collected data after 209 chunks in internlm2-math-plus-20b-IMat-GGUF/imatrix.dat |
|
|
|
llama_print_timings: load time = 7225.87 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 = 823077.43 ms / 107008 tokens ( 7.69 ms per token, 130.01 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 = 827539.28 ms / 107009 tokens |
|
|
|
Final estimate: PPL = 9.3305 +/- 0.11206 |
|
|
