eval/ppl_code_opti.log

0.04.879.314 I cmn          init: llama threadpool init, n_threads = 128
0.04.988.006 I 
0.04.988.226 I system_info: n_threads = 128 (n_threads_batch = 128) / 256 | CUDA : ARCHS = 860,890 | USE_GRAPHS = 1 | CPU : SSE3 = 1 | SSSE3 = 1 | AVX = 1 | AVX2 = 1 | F16C = 1 | FMA = 1 | BMI2 = 1 | LLAMAFILE = 1 | OPENMP = 1 | REPACK = 1 | 
0.04.988.245 I perplexity: tokenizing the input ..
0.05.501.249 I perplexity: tokenization took 512.991 ms
0.05.501.460 I perplexity: calculating perplexity over 96 chunks, n_ctx=1024, batch_size=2048, n_seq=2
0.08.294.568 I perplexity: 2.79 seconds per pass - ETA 2.23 minutes
[1]1.6110,[2]2.1845,[3]2.1495,[4]2.1638,[5]2.0233,[6]2.0755,[7]1.9536,[8]1.9023,[9]1.8552,[10]1.8262,[11]1.8365,[12]1.7954,[13]1.7659,[14]1.7480,[15]1.7198,[16]1.6960,[17]1.6918,[18]1.6541,[19]1.6280,[20]1.6278,[21]1.6344,[22]1.6359,[23]1.6308,[24]1.6298,[25]1.6161,[26]1.6177,[27]1.6217,[28]1.6171,[29]1.6350,[30]1.6255,[31]1.6357,[32]1.6337,[33]1.6340,[34]1.6384,[35]1.6408,[36]1.6393,[37]1.6511,[38]1.6480,[39]1.6421,[40]1.6600,[41]1.6723,[42]1.6768,[43]1.6747,[44]1.6633,[45]1.6721,[46]1.6930,[47]1.7102,[48]1.7257,[49]1.7212,[50]1.7238,[51]1.7376,[52]1.7581,[53]1.7504,[54]1.7562,[55]1.7519,[56]1.7538,[57]1.7542,[58]1.7527,[59]1.7576,[60]1.7675,[61]1.7787,[62]1.7825,[63]1.7892,[64]1.8053,[65]1.8241,[66]1.8595,[67]1.8585,[68]1.8757,[69]1.8785,[70]1.8760,[71]1.8759,[72]1.8698,[73]1.8690,[74]1.8736,[75]1.8778,[76]1.8699,[77]1.8625,[78]1.8572,[79]1.8515,[80]1.8461,[81]1.8453,[82]1.8610,[83]1.8622,[84]1.8722,[85]1.8703,[86]1.8674,[87]1.8721,[88]1.8719,[89]1.8673,[90]1.8726,[91]1.8725,[92]1.8723,[93]1.8835,[94]1.8865,[95]1.8805,[96]1.8773,
1.43.171.985 I Final estimate: PPL = 1.8773 +/- 0.01245