File size: 1,790 Bytes
27a5002 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 | 0.00.040.030 I common_init_result: fitting params to device memory ...
0.00.040.033 I common_init_result: (for bugs during this step try to reproduce them with -fit off, or provide --verbose logs if the bug only occurs with -fit on)
0.00.314.545 I common_params_fit_impl: projected to use 66756 MiB of host memory vs. 127438 MiB of total host memory
0.00.652.877 W llama_context: n_ctx_seq (2048) < n_ctx_train (262144) -- the full capacity of the model will not be utilized
0.00.672.925 W common_init_from_params: warming up the model with an empty run - please wait ... (--no-warmup to disable)
0.01.325.340 I
0.01.325.484 I system_info: n_threads = 16 (n_threads_batch = 16) / 32 | ROCm : NO_VMM = 1 | PEER_MAX_BATCH_SIZE = 128 | FA_ALL_QUANTS = 1 | CPU : SSE3 = 1 | SSSE3 = 1 | AVX = 1 | AVX_VNNI = 1 | AVX2 = 1 | F16C = 1 | FMA = 1 | BMI2 = 1 | AVX512 = 1 | AVX512_VBMI = 1 | AVX512_VNNI = 1 | AVX512_BF16 = 1 | LLAMAFILE = 1 | OPENMP = 1 | REPACK = 1 |
0.01.326.996 I perplexity: saving all logits to kld/bf16.kld
0.01.327.002 I perplexity: tokenizing the input ..
0.01.631.676 I perplexity: tokenization took 304.664 ms
0.01.631.766 I perplexity: calculating perplexity over 40 chunks, n_ctx=2048, batch_size=2048, n_seq=1
0.13.030.135 I perplexity: 11.26 seconds per pass - ETA 7.50 minutes
[1]5.6964,[2]6.6666,[3]7.0328,[4]7.2903,[5]7.1445,[6]6.1493,[7]5.7412,[8]5.6706,[9]5.9764,[10]6.0886,[11]6.1422,[12]6.3827,[13]6.4266,[14]6.4831,[15]6.5233,[16]6.6999,[17]6.7466,[18]6.8361,[19]6.7889,[20]6.5275,[21]6.5437,[22]6.5571,[23]6.6058,[24]6.6060,[25]6.6371,[26]6.6074,[27]6.7700,[28]6.8576,[29]6.8518,[30]6.7971,[31]6.6991,[32]6.6042,[33]6.5393,[34]6.5251,[35]6.5369,[36]6.5530,[37]6.4632,[38]6.3955,[39]6.3171,[40]6.2303,
7.32.185.106 I Final estimate: PPL = 6.2303 +/- 0.07538
|