@@ -6982,13 +6982,13 @@ static void llm_load_print_meta(llama_model_loader & ml, llama_model & model) {
69826982 // LLAMA_LOG_INFO("%s: model params = %.2f K\n", __func__, ml.n_elements*1e-3);
69836983 }
69846984
6985- LLAMA_LOG_INFO("%s: model size - %.2f Bytes (%.3f BPW) Total \n", __func__, ml.n_bytes/1.0, ml.n_bytes*8.0/ml.n_elements);
6986- LLAMA_LOG_INFO("%s: model size - %.2f KB (%.3f BPW) Total \n", __func__, ml.n_bytes/1000.0, ml.n_bytes*8.0/ml.n_elements);
6987- LLAMA_LOG_INFO("%s: model size - %.2f KiB (%.3f BPW) Total \n", __func__, ml.n_bytes/1024.0, ml.n_bytes*8.0/ml.n_elements);
6988- LLAMA_LOG_INFO("%s: model size - %.2f MB (%.3f BPW) Total \n", __func__, ml.n_bytes/1000.0/1000.0 , ml.n_bytes*8.0/ml.n_elements);
6989- LLAMA_LOG_INFO("%s: model size - %.2f MiB (%.3f BPW) Total \n", __func__, ml.n_bytes/1024.0/1024.0, ml.n_bytes*8.0/ml.n_elements);
6990- LLAMA_LOG_INFO("%s: model size - %.2f GB (%.3f BPW) Total \n", __func__, ml.n_bytes/1000.0/1000.0/1000.0, ml.n_bytes*8.0/ml.n_elements);
6991- LLAMA_LOG_INFO("%s: model size - %.2f GiB (%.3f BPW) Total \n", __func__, ml.n_bytes/1024.0/1024.0/1024.0, ml.n_bytes*8.0/ml.n_elements);
6985+ LLAMA_LOG_INFO("%s: model size = %.2f Bytes (%.3f BPW) \n", __func__, ml.n_bytes/1.0, ml.n_bytes*8.0/ml.n_elements);
6986+ LLAMA_LOG_INFO("%s: model size = %.2f KB (%.3f BPW) \n", __func__, ml.n_bytes/1000.0, ml.n_bytes*8.0/ml.n_elements);
6987+ LLAMA_LOG_INFO("%s: model size = %.2f KiB (%.3f BPW) \n", __func__, ml.n_bytes/1024.0, ml.n_bytes*8.0/ml.n_elements);
6988+ LLAMA_LOG_INFO("%s: model size = %.2f MB (%.3f BPW) \n", __func__, ml.n_bytes/1000.0/1000.0, ml.n_bytes*8.0/ml.n_elements);
6989+ LLAMA_LOG_INFO("%s: model size = %.2f MiB (%.3f BPW) \n", __func__, ml.n_bytes/1024.0/1024.0, ml.n_bytes*8.0/ml.n_elements);
6990+ LLAMA_LOG_INFO("%s: model size = %.2f GB (%.3f BPW) \n", __func__, ml.n_bytes/1000.0/1000.0/1000.0, ml.n_bytes*8.0/ml.n_elements);
6991+ LLAMA_LOG_INFO("%s: model size = %.2f GiB (%.3f BPW) \n", __func__, ml.n_bytes/1024.0/1024.0/1024.0, ml.n_bytes*8.0/ml.n_elements);
69926992
69936993 // if (ml.n_bytes < GiB) {
69946994 // LLAMA_LOG_INFO("%s: model size = %.3f MiB (%.3f BPW) \n", __func__, ml.n_bytes/1024.0/1024.0, ml.n_bytes*8.0/ml.n_elements);
@@ -7007,13 +7007,13 @@ static void llm_load_print_meta(llama_model_loader & ml, llama_model & model) {
70077007 n_bytes -= ggml_nbytes(meta_out);
70087008 n_elements -= ggml_nelements(meta_out);
70097009
7010- LLAMA_LOG_INFO("%s: repeating layers - %.2f Bytes (%.3f BPW) RL \n", __func__, n_bytes/1.0, n_bytes*8.0/n_elements);
7011- LLAMA_LOG_INFO("%s: repeating layers - %.2f KB (%.3f BPW) RL \n", __func__, n_bytes/1000.0, n_bytes*8.0/n_elements);
7012- LLAMA_LOG_INFO("%s: repeating layers - %.2f KiB (%.3f BPW) RL \n", __func__, n_bytes/1024.0, n_bytes*8.0/n_elements);
7013- LLAMA_LOG_INFO("%s: repeating layers - %.2f MB (%.3f BPW) RL \n", __func__, n_bytes/1000.0/1000.0, n_bytes*8.0/n_elements);
7014- LLAMA_LOG_INFO("%s: repeating layers - %.2f MiB (%.3f BPW) RL \n", __func__, n_bytes/1024.0/1024.0, n_bytes*8.0/n_elements);
7015- LLAMA_LOG_INFO("%s: repeating layers - %.2f GB (%.3f BPW) RL \n", __func__, n_bytes/1000.0/1000.0/1000.0, n_bytes*8.0/n_elements);
7016- LLAMA_LOG_INFO("%s: repeating layers - %.2f GiB (%.3f BPW) RL \n", __func__, n_bytes/1024.0/1024.0/1024.0, n_bytes*8.0/n_elements);
7010+ LLAMA_LOG_INFO("%s: repeating layers = %.2f Bytes (%.3f BPW) \n", __func__, n_bytes/1.0, n_bytes*8.0/n_elements);
7011+ LLAMA_LOG_INFO("%s: repeating layers = %.2f KB (%.3f BPW) \n", __func__, n_bytes/1000.0, n_bytes*8.0/n_elements);
7012+ LLAMA_LOG_INFO("%s: repeating layers = %.2f KiB (%.3f BPW) \n", __func__, n_bytes/1024.0, n_bytes*8.0/n_elements);
7013+ LLAMA_LOG_INFO("%s: repeating layers = %.2f MB (%.3f BPW) \n", __func__, n_bytes/1000.0/1000.0, n_bytes*8.0/n_elements);
7014+ LLAMA_LOG_INFO("%s: repeating layers = %.2f MiB (%.3f BPW) \n", __func__, n_bytes/1024.0/1024.0, n_bytes*8.0/n_elements);
7015+ LLAMA_LOG_INFO("%s: repeating layers = %.2f GB (%.3f BPW) \n", __func__, n_bytes/1000.0/1000.0/1000.0, n_bytes*8.0/n_elements);
7016+ LLAMA_LOG_INFO("%s: repeating layers = %.2f GiB (%.3f BPW) \n", __func__, n_bytes/1024.0/1024.0/1024.0, n_bytes*8.0/n_elements);
70177017
70187018 // if (n_bytes < GiB) {
70197019 // LLAMA_LOG_INFO("%s: repeating layers = %.3f MiB (%.3f BPW", __func__, n_bytes/1024.0/1024.0, n_bytes*8.0/n_elements);
0 commit comments