vrr/koboldcpp
2
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2026-10-02 19:15:48 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

remove some logspam (+1 squashed commits)

Browse source 
Squashed commits:

[c6f32e57e] remove some logspam
This commit is contained in:
Concedo 2026-09-27 01:23:46 +08:00
parent 2e7f67df40
commit 4959b8d369
2 changed files with 9 additions and 9 deletions
Download patch file Download diff file Expand all files Collapse all files

2
koboldcpp.py
View file

@ -81,7 +81,7 @@ dry_seq_break_max = 128
extra_images_max = 4 # for kontext/qwen img
# global vars
KcppVersion = "1.122"
KcppVersion = "1.122.1"
showdebug = True
kcpp_instance = None #global running instance
global_memory = {"tunnel_url": "", "restart_target":"", "input_to_exit":False, "load_complete":False, "restart_override_base_config":"", "last_active_timestamp":datetime.now(), "triggered_sleeping":False, "current_model":"initial_model", "base_config":"", "swapReqType": None, "loadedReqTypes": [], "autoswapmode": False}

16
src/llama-context.cpp
View file

@ -611,12 +611,12 @@ static int llama_graph_n_input_tensors(ggml_cgraph * gf) {
LLAMA_LOG_WARN("%s: input tensor '%32s' has op %s, expected GGML_OP_NONE\n",
__func__, tensor->name, ggml_op_name(tensor->op));
}
for (const ggml_tensor * node : nodes) {
LLAMA_LOG_DEBUG("%s: input tensor '%32s' [%s, ne = { %5" PRId64 ", %5" PRId64 ", %5" PRId64 ", %5" PRId64 " }] is used by node '%s' (%s)\n",
__func__, tensor->name, ggml_type_name(tensor->type),
tensor->ne[0], tensor->ne[1], tensor->ne[2], tensor->ne[3],
node->name, ggml_op_name(node->op));
}
// for (const ggml_tensor * node : nodes) {
// LLAMA_LOG_DEBUG("%s: input tensor '%32s' [%s, ne = { %5" PRId64 ", %5" PRId64 ", %5" PRId64 ", %5" PRId64 " }] is used by node '%s' (%s)\n",
// __func__, tensor->name, ggml_type_name(tensor->type),
// tensor->ne[0], tensor->ne[1], tensor->ne[2], tensor->ne[3],
// node->name, ggml_op_name(node->op));
// }
}
return (int) users.size();
@ -653,7 +653,7 @@ void llama_context::sched_reserve() {
if (memory) { //added by kcpp to reduce embeddings memory usage
llama_memory_context_ptr mctx;
if (memory) {
LLAMA_LOG_DEBUG("%s: reserving full memory module\n", __func__);
// LLAMA_LOG_DEBUG("%s: reserving full memory module\n", __func__);
mctx = memory->init_full();
if (!mctx) {
throw std::runtime_error("failed to initialize memory module");
@ -663,7 +663,7 @@ void llama_context::sched_reserve() {
// avoid reserving graphs with zero outputs - assume one output per sequence
const int n_outputs = n_seqs;
LLAMA_LOG_DEBUG("%s: worst-case: n_tokens = %d, n_seqs = %d, n_outputs = %d\n", __func__, n_tokens, n_seqs, n_outputs);
// LLAMA_LOG_DEBUG("%s: worst-case: n_tokens = %d, n_seqs = %d, n_outputs = %d\n", __func__, n_tokens, n_seqs, n_outputs);
resolve_fused_ops(mctx.get(), n_seqs);