File size: 2,497 Bytes
8686992 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 | 0.02.361.481 W model has unused tensor blk.64.attn_norm.weight (size = 20480 bytes) -- ignoring
0.02.361.496 W model has unused tensor blk.64.post_attention_norm.weight (size = 20480 bytes) -- ignoring
0.02.361.510 W model has unused tensor blk.64.attn_q.weight (size = 125829120 bytes) -- ignoring
0.02.361.515 W model has unused tensor blk.64.attn_k.weight (size = 10485760 bytes) -- ignoring
0.02.361.520 W model has unused tensor blk.64.attn_v.weight (size = 10485760 bytes) -- ignoring
0.02.361.537 W model has unused tensor blk.64.attn_output.weight (size = 62914560 bytes) -- ignoring
0.02.361.544 W model has unused tensor blk.64.attn_q_norm.weight (size = 1024 bytes) -- ignoring
0.02.361.549 W model has unused tensor blk.64.attn_k_norm.weight (size = 1024 bytes) -- ignoring
0.02.361.554 W model has unused tensor blk.64.ffn_gate.weight (size = 178257920 bytes) -- ignoring
0.02.361.561 W model has unused tensor blk.64.ffn_down.weight (size = 178257920 bytes) -- ignoring
0.02.361.567 W model has unused tensor blk.64.ffn_up.weight (size = 178257920 bytes) -- ignoring
0.02.361.575 W model has unused tensor blk.64.nextn.eh_proj.weight (size = 104857600 bytes) -- ignoring
0.02.361.583 W model has unused tensor blk.64.nextn.enorm.weight (size = 20480 bytes) -- ignoring
0.02.361.588 W model has unused tensor blk.64.nextn.hnorm.weight (size = 20480 bytes) -- ignoring
0.02.361.603 W model has unused tensor blk.64.nextn.shared_head_norm.weight (size = 20480 bytes) -- ignoring
0.11.979.495 I cmn init: llama threadpool init, n_threads = 64
0.13.117.567 I
0.13.117.843 I system_info: n_threads = 64 (n_threads_batch = 64) / 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 |
0.13.117.876 I perplexity: tokenizing the input ..
0.13.622.600 I perplexity: tokenization took 504.706 ms
0.13.622.862 I perplexity: calculating perplexity over 32 chunks, n_ctx=1024, batch_size=512, n_seq=1
0.20.701.304 I perplexity: 7.08 seconds per pass - ETA 3.77 minutes
[1]1.1416,[2]1.5868,[3]1.6754,[4]1.7163,[5]1.6527,[6]1.7463,[7]1.6782,[8]1.6518,[9]1.6303,[10]1.6124,[11]1.5941,[12]1.5558,[13]1.5194,[14]1.4940,[15]1.4682,[16]1.4478,[17]1.4363,[18]1.4100,[19]1.3935,[20]1.3923,[21]1.3870,[22]1.3866,[23]1.3765,[24]1.3758,[25]1.3632,[26]1.3524,[27]1.3429,[28]1.3295,[29]1.3431,[30]1.3346,[31]1.3386,[32]1.3417,
3.37.343.457 I Final estimate: PPL = 1.3417 +/- 0.01215
|