upstream-archive/stable-diffusion.cpp
1
0
Fork 0
mirror of https://github.com/leejet/stable-diffusion.cpp.git synced 2026-10-03 03:17:51 -05:00
Code Issues Packages Projects Releases Wiki Activity

feat: add verbose logging and log-level selection (#1941)

Browse Source
This commit is contained in:
leejet
2026-09-06 23:14:03 +08:00
committed by GitHub
parent 462d675018
commit dbb611264e
69 changed files with 551 additions and 484 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/model/adapter/lora.hpp
+3 -3
View File
@@ -120,7 +120,7 @@ struct LoraModel : public GGMLRunner {
return false;
}
LOG_DEBUG("finished loaded lora");
LOG_VERBOSE("finished loaded lora");
return true;
}
@@ -242,7 +242,7 @@ struct LoraModel : public GGMLRunner {
if (iter != lora_tensors.end()) {
float alpha = ggml_ext_backend_tensor_get_f32(iter->second);
scale_value = alpha / rank;
// LOG_DEBUG("rank %s %ld %.2f %.2f", alpha_name.c_str(), rank, alpha, scale_value);
// LOG_VERBOSE("rank %s %ld %.2f %.2f", alpha_name.c_str(), rank, alpha, scale_value);
applied_lora_tensors.insert(alpha_name);
}
}
@@ -798,7 +798,7 @@ struct LoraModel : public GGMLRunner {
float alpha = ggml_ext_backend_tensor_get_f32(iter->second);
scale_value = alpha / rank;
scale_tensor_name = alpha_name;
// LOG_DEBUG("rank %s %ld %.2f %.2f", alpha_name.c_str(), rank, alpha, scale_value);
// LOG_VERBOSE("rank %s %ld %.2f %.2f", alpha_name.c_str(), rank, alpha, scale_value);
}
}
scale_value *= multiplier;
src/model/adapter/pmid.hpp
+1 -1
View File
@@ -639,7 +639,7 @@ struct PhotoMakerIDEmbed : public GGMLRunner {
return false;
}
LOG_DEBUG("finished loading PhotoMaker ID Embeds ");
LOG_VERBOSE("finished loading PhotoMaker ID Embeds ");
return true;
}
src/model/common/block.hpp
+2 -2
View File
@@ -340,7 +340,7 @@ public:
enable_ip(enable_ip) {
int64_t inner_dim = d_head * n_head;
if (context_dim == 320 && d_head == 320) {
// LOG_DEBUG("CrossAttention: temp set dim to 1024 for sdxs_09");
// LOG_VERBOSE("CrossAttention: temp set dim to 1024 for sdxs_09");
xtra_dim = true;
context_dim = 1024;
}
@@ -370,7 +370,7 @@ public:
auto q = to_q->forward(ctx, x); // [N, n_token, inner_dim]
if (xtra_dim) {
// LOG_DEBUG("CrossAttention: temp set dim to 1024 for sdxs_09");
// LOG_VERBOSE("CrossAttention: temp set dim to 1024 for sdxs_09");
context->ne[0] = 1024; // patch dim
}
auto k = to_k->forward(ctx, context); // [N, n_context, inner_dim]
src/model/detector/yolov8.h
+6 -6
View File
@@ -68,12 +68,12 @@ struct YOLOv8Config {
}
if (config.valid) {
LOG_DEBUG("yolov8: classes=%d, reg_max=%d, p3=%d, p4=%d, p5=%d",
config.num_classes,
config.reg_max,
config.out_channels[15],
config.out_channels[18],
config.out_channels[21]);
LOG_VERBOSE("yolov8: classes=%d, reg_max=%d, p3=%d, p4=%d, p5=%d",
config.num_classes,
config.reg_max,
config.out_channels[15],
config.out_channels[18],
config.out_channels[21]);
}
return config;
}
src/model/diffusion/anima.hpp
+5 -5
View File
@@ -46,11 +46,11 @@ namespace Anima {
}
if (detected_layers > 0) {
config.num_layers = detected_layers;
LOG_DEBUG("anima: num_layers = %" PRId64 ", hidden_size = %" PRId64 ", num_heads = %" PRId64 ", head_dim = %" PRId64,
config.num_layers,
config.hidden_size,
config.num_heads,
config.head_dim);
LOG_VERBOSE("anima: num_layers = %" PRId64 ", hidden_size = %" PRId64 ", num_heads = %" PRId64 ", head_dim = %" PRId64,
config.num_layers,
config.hidden_size,
config.num_heads,
config.head_dim);
}
return config;
}
src/model/diffusion/boogu.hpp
+10 -10
View File
@@ -109,16 +109,16 @@ namespace Boogu {
}
config.timestep_embed_dim = std::min<int64_t>(config.hidden_size, 1024);
LOG_DEBUG("boogu_image: layers=%" PRId64 ", double_stream_layers=%" PRId64 ", refiner_layers=%" PRId64 ", hidden=%" PRId64 ", heads=%" PRId64 ", kv_heads=%" PRId64 ", head_dim=%" PRId64 ", in_channels=%" PRId64 ", out_channels=%" PRId64,
config.num_layers,
config.num_double_stream_layers,
config.num_refiner_layers,
config.hidden_size,
config.num_attention_heads,
config.num_kv_heads,
config.head_dim,
config.in_channels,
config.out_channels);
LOG_VERBOSE("boogu_image: layers=%" PRId64 ", double_stream_layers=%" PRId64 ", refiner_layers=%" PRId64 ", hidden=%" PRId64 ", heads=%" PRId64 ", kv_heads=%" PRId64 ", head_dim=%" PRId64 ", in_channels=%" PRId64 ", out_channels=%" PRId64,
config.num_layers,
config.num_double_stream_layers,
config.num_refiner_layers,
config.hidden_size,
config.num_attention_heads,
config.num_kv_heads,
config.head_dim,
config.in_channels,
config.out_channels);
return config;
}
};
src/model/diffusion/ernie_image.hpp
+7 -7
View File
@@ -72,13 +72,13 @@ namespace ErnieImage {
for (int axis_dim : config.axes_dim) {
config.axes_dim_sum += axis_dim;
}
LOG_DEBUG("ernie_image: num_layers = %" PRId64 ", hidden_size = %" PRId64 ", num_heads = %" PRId64 ", ffn_hidden_size = %" PRId64 ", in_channels = %" PRId64 ", out_channels = %" PRId64,
config.num_layers,
config.hidden_size,
config.num_heads,
config.ffn_hidden_size,
config.in_channels,
config.out_channels);
LOG_VERBOSE("ernie_image: num_layers = %" PRId64 ", hidden_size = %" PRId64 ", num_heads = %" PRId64 ", ffn_hidden_size = %" PRId64 ", in_channels = %" PRId64 ", out_channels = %" PRId64,
config.num_layers,
config.hidden_size,
config.num_heads,
config.ffn_hidden_size,
config.in_channels,
config.out_channels);
return config;
}
};
src/model/diffusion/flux.hpp
+13 -13
View File
@@ -123,16 +123,16 @@ namespace Flux {
config.guidance_embed = true;
}
if (name.find("__x0__") != std::string::npos) {
LOG_DEBUG("using x0 prediction");
LOG_VERBOSE("using x0 prediction");
config.chroma_radiance_params.use_x0 = true;
}
if (name.find("__32x32__") != std::string::npos) {
LOG_DEBUG("using patch size 32");
LOG_VERBOSE("using patch size 32");
config.patch_size = 32;
}
if (name.find("img_in_patch.weight") != std::string::npos) {
actual_radiance_patch_size = tensor_storage.ne[0];
LOG_DEBUG("actual radiance patch size: %" PRId64, actual_radiance_patch_size);
LOG_VERBOSE("actual radiance patch size: %" PRId64, actual_radiance_patch_size);
}
if (name.find("distilled_guidance_layer.in_proj.weight") != std::string::npos) {
config.is_chroma = true;
@@ -169,7 +169,7 @@ namespace Flux {
}
if (actual_radiance_patch_size > 0 && actual_radiance_patch_size != config.patch_size) {
GGML_ASSERT(config.patch_size == 2 * actual_radiance_patch_size);
LOG_DEBUG("using fake x2 patch size");
LOG_VERBOSE("using fake x2 patch size");
config.chroma_radiance_params.fake_patch_size_x2 = true;
}
if (head_dim > 0) {
@@ -179,13 +179,13 @@ namespace Flux {
for (int axis_dim : config.axes_dim) {
config.axes_dim_sum += axis_dim;
}
LOG_DEBUG("flux: depth = %d, depth_single_blocks = %d, guidance_embed = %s, context_in_dim = %" PRId64 ", hidden_size = %" PRId64 ", num_heads = %d",
config.depth,
config.depth_single_blocks,
config.guidance_embed ? "true" : "false",
config.context_in_dim,
config.hidden_size,
config.num_heads);
LOG_VERBOSE("flux: depth = %d, depth_single_blocks = %d, guidance_embed = %s, context_in_dim = %" PRId64 ", hidden_size = %" PRId64 ", num_heads = %d",
config.depth,
config.depth_single_blocks,
config.guidance_embed ? "true" : "false",
config.context_in_dim,
config.hidden_size,
config.num_heads);
return config;
}
};
@@ -1560,7 +1560,7 @@ namespace Flux {
config.axes_dim,
sd_version_is_longcat(version));
int pos_len = static_cast<int>(pe_vec.size() / config.axes_dim_sum / 2);
// LOG_DEBUG("pos_len %d", pos_len);
// LOG_VERBOSE("pos_len %d", pos_len);
auto pe = ggml_new_tensor_4d(compute_ctx, GGML_TYPE_F32, 2, 2, config.axes_dim_sum / 2, pos_len);
// pe->data = pe_vec.data();
// print_ggml_tensor(pe);
@@ -1702,7 +1702,7 @@ namespace Flux {
GGML_ASSERT(!out_opt.empty());
out = std::move(out_opt);
print_sd_tensor(out);
LOG_DEBUG("flux test done in %lldms", t1 - t0);
LOG_VERBOSE("flux test done in %lldms", t1 - t0);
}
}
src/model/diffusion/hunyuan.hpp
+11 -11
View File
@@ -266,16 +266,16 @@ namespace Hunyuan {
GGML_ASSERT(config.hidden_size / config.num_heads == config.axes_dim_sum);
if (inferred) {
LOG_DEBUG("hunyuan video: depth = %d, single depth = %d, in_channels = %" PRId64 ", out_channels = %" PRId64 ", hidden_size = %" PRId64 ", context_in_dim = %" PRId64 ", patch_size = %dx%dx%d",
config.depth,
config.depth_single_blocks,
config.in_channels,
config.out_channels,
config.hidden_size,
config.context_in_dim,
std::get<0>(config.patch_size),
std::get<1>(config.patch_size),
std::get<2>(config.patch_size));
LOG_VERBOSE("hunyuan video: depth = %d, single depth = %d, in_channels = %" PRId64 ", out_channels = %" PRId64 ", hidden_size = %" PRId64 ", context_in_dim = %" PRId64 ", patch_size = %dx%dx%d",
config.depth,
config.depth_single_blocks,
config.in_channels,
config.out_channels,
config.hidden_size,
config.context_in_dim,
std::get<0>(config.patch_size),
std::get<1>(config.patch_size),
std::get<2>(config.patch_size));
}
return config;
}
@@ -615,7 +615,7 @@ namespace Hunyuan {
config.theta,
config.axes_dim);
int64_t pos_len = static_cast<int64_t>(pe_vec.size() / config.axes_dim_sum / 2);
// LOG_DEBUG("pos_len %d", pos_len);
// LOG_VERBOSE("pos_len %d", pos_len);
auto pe = ggml_new_tensor_4d(compute_ctx, GGML_TYPE_F32, 2, 2, config.axes_dim_sum / 2, pos_len);
// pe->data = pe_vec.data();
// print_ggml_tensor(pe, true, "pe");
src/model/diffusion/ideogram4.hpp
+6 -6
View File
@@ -58,11 +58,11 @@ namespace Ideogram4 {
}
if (detected_layers > 0) {
config.num_layers = detected_layers;
LOG_DEBUG("ideogram4: num_layers = %" PRId64 ", emb_dim = %" PRId64 ", num_heads = %" PRId64 ", intermediate_size = %" PRId64,
config.num_layers,
config.emb_dim,
config.num_heads,
config.intermediate_size);
LOG_VERBOSE("ideogram4: num_layers = %" PRId64 ", emb_dim = %" PRId64 ", num_heads = %" PRId64 ", intermediate_size = %" PRId64,
config.num_layers,
config.emb_dim,
config.num_heads,
config.intermediate_size);
}
return config;
}
@@ -465,7 +465,7 @@ namespace Ideogram4 {
}
}
if (has_uncond_model) {
LOG_DEBUG("using uncond model");
LOG_VERBOSE("using uncond model");
uncond_model = Ideogram4Transformer(config);
uncond_model.init(params_ctx, tensor_storage_map, uncond_prefix);
}
src/model/diffusion/krea2.hpp
+10 -10
View File
@@ -143,16 +143,16 @@ namespace Krea2 {
}
config.update_axes_dim();
LOG_DEBUG("krea2: layers=%" PRId64 ", features=%" PRId64 ", heads=%" PRId64 ", kv_heads=%" PRId64 ", text_dim=%" PRId64 ", text_layers=%" PRId64 ", text_heads=%" PRId64 ", text_kv_heads=%" PRId64 ", channels=%" PRId64,
config.layers,
config.features,
config.heads,
config.kv_heads,
config.text_dim,
config.text_layers,
config.text_heads,
config.text_kv_heads,
config.in_channels);
LOG_VERBOSE("krea2: layers=%" PRId64 ", features=%" PRId64 ", heads=%" PRId64 ", kv_heads=%" PRId64 ", text_dim=%" PRId64 ", text_layers=%" PRId64 ", text_heads=%" PRId64 ", text_kv_heads=%" PRId64 ", channels=%" PRId64,
config.layers,
config.features,
config.heads,
config.kv_heads,
config.text_dim,
config.text_layers,
config.text_heads,
config.text_kv_heads,
config.in_channels);
return config;
}
};
src/model/diffusion/lens.hpp
+8 -8
View File
@@ -66,14 +66,14 @@ namespace Lens {
for (int axis_dim : config.axes_dim) {
config.axes_dim_sum += axis_dim;
}
LOG_DEBUG("lens: num_layers = %d, selected_layer_count = %d, hidden_size = %" PRId64 ", num_attention_heads = %" PRId64 ", attention_head_dim = %" PRId64 ", in_channels = %" PRId64 ", out_channels = %" PRId64,
config.num_layers,
config.selected_layer_count,
config.num_attention_heads * config.attention_head_dim,
config.num_attention_heads,
config.attention_head_dim,
config.in_channels,
config.out_channels);
LOG_VERBOSE("lens: num_layers = %d, selected_layer_count = %d, hidden_size = %" PRId64 ", num_attention_heads = %" PRId64 ", attention_head_dim = %" PRId64 ", in_channels = %" PRId64 ", out_channels = %" PRId64,
config.num_layers,
config.selected_layer_count,
config.num_attention_heads * config.attention_head_dim,
config.num_attention_heads,
config.attention_head_dim,
config.in_channels,
config.out_channels);
return config;
}
};
src/model/diffusion/lingbot_video.hpp
+11 -11
View File
@@ -127,17 +127,17 @@ namespace LingBotVideo {
config.topk_group = 2;
config.routed_scaling_factor = 2.5f;
}
LOG_DEBUG("lingbot_video: depth = %" PRId64 ", hidden_size = %" PRId64 ", heads = %" PRId64 ", text_dim = %" PRId64 ", experts = %" PRId64 ", experts_per_tok = %" PRId64 ", n_group = %" PRId64 ", topk_group = %" PRId64 ", route_scale = %.2f, sparse_layers = %zu",
config.depth,
config.hidden_size,
config.num_attention_heads,
config.text_dim,
config.num_experts,
config.num_experts_per_tok,
config.n_group,
config.topk_group,
config.routed_scaling_factor,
config.sparse_layers.size());
LOG_VERBOSE("lingbot_video: depth = %" PRId64 ", hidden_size = %" PRId64 ", heads = %" PRId64 ", text_dim = %" PRId64 ", experts = %" PRId64 ", experts_per_tok = %" PRId64 ", n_group = %" PRId64 ", topk_group = %" PRId64 ", route_scale = %.2f, sparse_layers = %zu",
config.depth,
config.hidden_size,
config.num_attention_heads,
config.text_dim,
config.num_experts,
config.num_experts_per_tok,
config.n_group,
config.topk_group,
config.routed_scaling_factor,
config.sparse_layers.size());
return config;
}
};
src/model/diffusion/ltxv.hpp
+7 -7
View File
@@ -274,12 +274,12 @@ namespace LTXV {
config.audio_connector_apply_gated_attention = true;
}
}
LOG_DEBUG("ltxav: num_layers = %" PRId64 ", hidden_size = %" PRId64 ", num_attention_heads = %" PRId64 ", audio_hidden_size = %" PRId64 ", audio_num_attention_heads = %" PRId64,
config.num_layers,
config.hidden_size,
config.num_attention_heads,
config.audio_hidden_size,
config.audio_num_attention_heads);
LOG_VERBOSE("ltxav: num_layers = %" PRId64 ", hidden_size = %" PRId64 ", num_attention_heads = %" PRId64 ", audio_hidden_size = %" PRId64 ", audio_num_attention_heads = %" PRId64,
config.num_layers,
config.hidden_size,
config.num_attention_heads,
config.audio_hidden_size,
config.audio_num_attention_heads);
return config;
}
};
@@ -2070,7 +2070,7 @@ namespace LTXV {
GGML_ASSERT(!out_opt.empty());
print_sd_tensor(out_opt, false, "ltxav_out");
LOG_DEBUG("ltxav test done in %lldms", t1 - t0);
LOG_VERBOSE("ltxav test done in %lldms", t1 - t0);
}
static void load_from_file_and_test(const std::string& model_path,
src/model/diffusion/minimax_h3.hpp
+8 -8
View File
@@ -106,14 +106,14 @@ namespace MiniMaxH3 {
config.rope_inv_freq_len = inv_freq->ne[0];
}
LOG_DEBUG("minimax_h3: layers=%" PRId64 ", hidden=%" PRId64 ", heads=%" PRId64
", head_dim=%" PRId64 ", ffn=%" PRId64 ", adaln_curve=%" PRId64,
config.num_layers,
config.hidden_size,
config.num_attention_heads,
config.attention_head_dim,
config.ffn_hidden_size,
config.adaln_curve_grid);
LOG_VERBOSE("minimax_h3: layers=%" PRId64 ", hidden=%" PRId64 ", heads=%" PRId64
", head_dim=%" PRId64 ", ffn=%" PRId64 ", adaln_curve=%" PRId64,
config.num_layers,
config.hidden_size,
config.num_attention_heads,
config.attention_head_dim,
config.ffn_hidden_size,
config.adaln_curve_grid);
return config;
}
};
src/model/diffusion/minit2i.hpp
+9 -9
View File
@@ -108,15 +108,15 @@ namespace MiniT2I {
config.head_dim = config.hidden_size == 1248 ? 52 : 64;
config.num_heads = config.hidden_size / config.head_dim;
}
LOG_DEBUG("minit2i: hidden_size=%" PRId64 ", txt_hidden_size=%" PRId64 ", heads=%" PRId64 ", head_dim=%" PRId64 ", double_blocks=%" PRId64 ", txt_blocks=%" PRId64 ", patch=%" PRId64 ", in_channels=%" PRId64,
config.hidden_size,
config.txt_hidden_size,
config.num_heads,
config.head_dim,
config.depth_double,
config.txt_preamble_depth,
config.patch_size,
config.in_channels);
LOG_VERBOSE("minit2i: hidden_size=%" PRId64 ", txt_hidden_size=%" PRId64 ", heads=%" PRId64 ", head_dim=%" PRId64 ", double_blocks=%" PRId64 ", txt_blocks=%" PRId64 ", patch=%" PRId64 ", in_channels=%" PRId64,
config.hidden_size,
config.txt_hidden_size,
config.num_heads,
config.head_dim,
config.depth_double,
config.txt_preamble_depth,
config.patch_size,
config.in_channels);
return config;
}
};
src/model/diffusion/mmdit.hpp
+11 -11
View File
@@ -120,16 +120,16 @@ struct MMDiTConfig {
}
if (has_weight_config) {
LOG_DEBUG("mmdit: num_layers = %" PRId64 ", num_mmdit_x_layers = %" PRId64 ", hidden_size = %" PRId64 ", patch_size = %d, in_channels = %" PRId64 ", out_channels = %" PRId64 ", context_size = %" PRId64 ", adm_in_channels = %" PRId64 ", qk_norm = %s",
config.depth,
config.d_self + 1,
config.hidden_size,
config.patch_size,
config.in_channels,
config.out_channels,
config.context_size,
config.adm_in_channels,
config.qk_norm.empty() ? "none" : config.qk_norm.c_str());
LOG_VERBOSE("mmdit: num_layers = %" PRId64 ", num_mmdit_x_layers = %" PRId64 ", hidden_size = %" PRId64 ", patch_size = %d, in_channels = %" PRId64 ", out_channels = %" PRId64 ", context_size = %" PRId64 ", adm_in_channels = %" PRId64 ", qk_norm = %s",
config.depth,
config.d_self + 1,
config.hidden_size,
config.patch_size,
config.in_channels,
config.out_channels,
config.context_size,
config.adm_in_channels,
config.qk_norm.empty() ? "none" : config.qk_norm.c_str());
}
return config;
}
@@ -1045,7 +1045,7 @@ struct MMDiTRunner : public DiffusionModelRunner {
GGML_ASSERT(!out_opt.empty());
out = std::move(out_opt);
print_sd_tensor(out);
LOG_DEBUG("mmdit test done in %lldms", t1 - t0);
LOG_VERBOSE("mmdit test done in %lldms", t1 - t0);
}
}
src/model/diffusion/pid.hpp
+10 -10
View File
@@ -109,16 +109,16 @@ namespace Pid {
config.lq_latent_channels = latent_proj_in_channels;
config.lq_latent_down_factor = latent_proj_in_channels >= 64 ? 16 : 8;
}
LOG_DEBUG("pid: version = %s, patch_depth = %" PRId64 ", pixel_depth = %" PRId64 ", patch_mlp_hidden_dim = %" PRId64 ", lq_latent_channels = %" PRId64 ", lq_hidden_dim = %" PRId64 ", lq_latent_down_factor = %" PRId64 ", lq_latent_unpatchify_factor = %" PRId64 ", lq_interval = %" PRId64,
config.pit_lq_inject ? "1.5" : "1",
config.patch_depth,
config.pixel_depth,
config.patch_mlp_hidden_dim,
config.lq_latent_channels,
config.lq_hidden_dim,
config.lq_latent_down_factor,
config.lq_latent_unpatchify_factor,
config.lq_interval);
LOG_VERBOSE("pid: version = %s, patch_depth = %" PRId64 ", pixel_depth = %" PRId64 ", patch_mlp_hidden_dim = %" PRId64 ", lq_latent_channels = %" PRId64 ", lq_hidden_dim = %" PRId64 ", lq_latent_down_factor = %" PRId64 ", lq_latent_unpatchify_factor = %" PRId64 ", lq_interval = %" PRId64,
config.pit_lq_inject ? "1.5" : "1",
config.patch_depth,
config.pixel_depth,
config.patch_mlp_hidden_dim,
config.lq_latent_channels,
config.lq_hidden_dim,
config.lq_latent_down_factor,
config.lq_latent_unpatchify_factor,
config.lq_interval);
return config;
}
};
src/model/diffusion/qwen_image.hpp
+5 -5
View File
@@ -49,9 +49,9 @@ namespace Qwen {
}
}
}
LOG_DEBUG("qwen_image: num_layers = %d, zero_cond_t = %s",
config.num_layers,
config.zero_cond_t ? "true" : "false");
LOG_VERBOSE("qwen_image: num_layers = %d, zero_cond_t = %s",
config.num_layers,
config.zero_cond_t ? "true" : "false");
return config;
}
};
@@ -646,7 +646,7 @@ namespace Qwen {
circular_x_enabled,
config.axes_dim);
int pos_len = static_cast<int>(pe_vec.size() / config.axes_dim_sum / 2);
// LOG_DEBUG("pos_len %d", pos_len);
// LOG_VERBOSE("pos_len %d", pos_len);
auto pe = ggml_new_tensor_4d(compute_ctx, GGML_TYPE_F32, 2, 2, config.axes_dim_sum / 2, pos_len);
// pe->data = pe_vec.data();
// print_ggml_tensor(pe, true, "pe");
@@ -760,7 +760,7 @@ namespace Qwen {
GGML_ASSERT(!out_opt.empty());
out = std::move(out_opt);
print_sd_tensor(out);
LOG_DEBUG("qwen_image test done in %lldms", t1 - t0);
LOG_VERBOSE("qwen_image test done in %lldms", t1 - t0);
}
}
src/model/diffusion/sefi_image.hpp
+4 -4
View File
@@ -34,10 +34,10 @@ namespace SefiImage {
config.hidden_size = tensor_storage.ne[1] * 2;
}
}
LOG_DEBUG("sefi_image: semantic_channels = %" PRId64 ", texture_latent_channels = %" PRId64 ", hidden_size = %" PRId64,
config.semantic_channels,
config.texture_latent_channels,
config.hidden_size);
LOG_VERBOSE("sefi_image: semantic_channels = %" PRId64 ", texture_latent_channels = %" PRId64 ", hidden_size = %" PRId64,
config.semantic_channels,
config.texture_latent_channels,
config.hidden_size);
return config;
}
};
src/model/diffusion/unet.hpp
+10 -10
View File
@@ -128,15 +128,15 @@ struct UNetConfig {
}
}
LOG_DEBUG("unet: in_channels = %d, out_channels = %d, model_channels = %d, time_embed_dim = %d, context_dim = %d, adm_in_channels = %d, num_res_blocks = %d, tiny_unet = %s",
config.in_channels,
config.out_channels,
config.model_channels,
config.time_embed_dim,
config.context_dim,
config.adm_in_channels,
config.num_res_blocks,
config.tiny_unet ? "true" : "false");
LOG_VERBOSE("unet: in_channels = %d, out_channels = %d, model_channels = %d, time_embed_dim = %d, context_dim = %d, adm_in_channels = %d, num_res_blocks = %d, tiny_unet = %s",
config.in_channels,
config.out_channels,
config.model_channels,
config.time_embed_dim,
config.context_dim,
config.adm_in_channels,
config.num_res_blocks,
config.tiny_unet ? "true" : "false");
return config;
}
};
@@ -904,7 +904,7 @@ struct UNetModelRunner : public DiffusionModelRunner {
GGML_ASSERT(!out_opt.empty());
out = std::move(out_opt);
print_sd_tensor(out);
LOG_DEBUG("unet test done in %lldms", t1 - t0);
LOG_VERBOSE("unet test done in %lldms", t1 - t0);
}
}
};
src/model/diffusion/wan.hpp
+9 -9
View File
@@ -75,13 +75,13 @@ namespace WAN {
config.flf_pos_embed_token_number = 514;
}
}
LOG_DEBUG("wan: model_type = %s, num_layers = %d, vace_layers = %d, dim = %" PRId64 ", ffn_dim = %" PRId64 ", num_heads = %" PRId64,
config.model_type.c_str(),
config.num_layers,
config.vace_layers,
config.dim,
config.ffn_dim,
config.num_heads);
LOG_VERBOSE("wan: model_type = %s, num_layers = %d, vace_layers = %d, dim = %" PRId64 ", ffn_dim = %" PRId64 ", num_heads = %" PRId64,
config.model_type.c_str(),
config.num_layers,
config.vace_layers,
config.dim,
config.ffn_dim,
config.num_heads);
return config;
}
};
@@ -909,7 +909,7 @@ namespace WAN {
config.theta,
config.axes_dim);
int pos_len = static_cast<int>(pe_vec.size() / config.axes_dim_sum / 2);
// LOG_DEBUG("pos_len %d", pos_len);
// LOG_VERBOSE("pos_len %d", pos_len);
auto pe = ggml_new_tensor_4d(compute_ctx, GGML_TYPE_F32, 2, 2, config.axes_dim_sum / 2, pos_len);
// pe->data = pe_vec.data();
// print_ggml_tensor(pe);
@@ -1007,7 +1007,7 @@ namespace WAN {
GGML_ASSERT(!out_opt.empty());
out = std::move(out_opt);
print_sd_tensor(out);
LOG_DEBUG("wan test done in %lldms", t1 - t0);
LOG_VERBOSE("wan test done in %lldms", t1 - t0);
}
}
src/model/diffusion/z_image.hpp
+10 -10
View File
@@ -107,14 +107,14 @@ namespace ZImage {
config.num_kv_heads = std::max<int64_t>(1, (qkv_heads - config.num_heads) / 2);
}
}
LOG_DEBUG("z_image: num_layers = %" PRId64 ", num_refiner_layers = %" PRId64 ", hidden_size = %" PRId64 ", num_heads = %" PRId64 ", num_kv_heads = %" PRId64 ", in_channels = %" PRId64 ", out_channels = %" PRId64,
config.num_layers,
config.num_refiner_layers,
config.hidden_size,
config.num_heads,
config.num_kv_heads,
config.in_channels,
config.out_channels);
LOG_VERBOSE("z_image: num_layers = %" PRId64 ", num_refiner_layers = %" PRId64 ", hidden_size = %" PRId64 ", num_heads = %" PRId64 ", num_kv_heads = %" PRId64 ", in_channels = %" PRId64 ", out_channels = %" PRId64,
config.num_layers,
config.num_refiner_layers,
config.hidden_size,
config.num_heads,
config.num_kv_heads,
config.in_channels,
config.out_channels);
return config;
}
};
@@ -603,7 +603,7 @@ namespace ZImage {
circular_x_enabled,
config.axes_dim);
int pos_len = static_cast<int>(pe_vec.size() / config.axes_dim_sum / 2);
// LOG_DEBUG("pos_len %d", pos_len);
// LOG_VERBOSE("pos_len %d", pos_len);
auto pe = ggml_new_tensor_4d(compute_ctx, GGML_TYPE_F32, 2, 2, config.axes_dim_sum / 2, pos_len);
// pe->data = pe_vec.data();
// print_ggml_tensor(pe, true, "pe");
@@ -689,7 +689,7 @@ namespace ZImage {
GGML_ASSERT(!out_opt.empty());
out = std::move(out_opt);
print_sd_tensor(out);
LOG_DEBUG("z_image test done in %lldms", t1 - t0);
LOG_VERBOSE("z_image test done in %lldms", t1 - t0);
}
}
src/model/te/clip.hpp
+4 -4
View File
@@ -100,13 +100,13 @@ public:
const std::string& graph_cut_prefix = "") {
// x: [N, n_token, d_model]
int layer_idx = n_layer - 1;
// LOG_DEBUG("clip_skip %d", clip_skip);
// LOG_VERBOSE("clip_skip %d", clip_skip);
if (clip_skip > 0) {
layer_idx = n_layer - clip_skip;
}
for (int i = 0; i < n_layer; i++) {
// LOG_DEBUG("layer %d", i);
// LOG_VERBOSE("layer %d", i);
if (i == layer_idx + 1) {
break;
}
@@ -116,7 +116,7 @@ public:
if (!graph_cut_prefix.empty()) {
sd::ggml_graph_cut::mark_graph_cut(x, graph_cut_prefix + ".layers." + std::to_string(i), "x");
}
// LOG_DEBUG("layer %d", i);
// LOG_VERBOSE("layer %d", i);
}
return x;
}
@@ -320,7 +320,7 @@ public:
if (text_projection != nullptr) {
pooled = ggml_ext_linear(ctx->ggml_ctx, pooled, text_projection, nullptr);
} else {
LOG_DEBUG("identity projection");
LOG_VERBOSE("identity projection");
}
return pooled; // [hidden_size, 1, 1]
}
src/model/te/llm.hpp
+14 -14
View File
@@ -319,11 +319,11 @@ namespace LLM {
config.vision.deepstack_visual_indexes = {8, 16, 24};
}
}
LOG_DEBUG("llm: num_layers = %" PRId64 ", vocab_size = %" PRId64 ", hidden_size = %" PRId64 ", intermediate_size = %" PRId64,
config.num_layers,
config.vocab_size,
config.hidden_size,
config.intermediate_size);
LOG_VERBOSE("llm: num_layers = %" PRId64 ", vocab_size = %" PRId64 ", hidden_size = %" PRId64 ", intermediate_size = %" PRId64,
config.num_layers,
config.vocab_size,
config.hidden_size,
config.intermediate_size);
return config;
}
};
@@ -1887,9 +1887,9 @@ namespace LLM {
enable_vision = false;
}
if (enable_vision) {
LOG_DEBUG("enable llm vision");
LOG_VERBOSE("enable llm vision");
if (config.llama_cpp_style) {
LOG_DEBUG("llama.cpp style vision weight");
LOG_VERBOSE("llama.cpp style vision weight");
}
}
model = LLM(config, enable_vision, config.llama_cpp_style);
@@ -2375,7 +2375,7 @@ namespace LLM {
ss << "['" << item.first << "', " << item.second << "], ";
}
ss << "]";
LOG_DEBUG("parse '%s' to %s", text.c_str(), ss.str().c_str());
LOG_VERBOSE("parse '%s' to %s", text.c_str(), ss.str().c_str());
}
std::vector<int> tokens;
@@ -2426,7 +2426,7 @@ namespace LLM {
out = std::move(out_opt);
print_sd_tensor(out, false, "image_embed");
image_embed = out;
LOG_DEBUG("llm encode_image test done in %lldms", t1 - t0);
LOG_VERBOSE("llm encode_image test done in %lldms", t1 - t0);
}
std::string placeholder = "<|image_pad|>";
@@ -2466,7 +2466,7 @@ namespace LLM {
GGML_ASSERT(!out_opt.empty());
out = std::move(out_opt);
print_sd_tensor(out);
LOG_DEBUG("llm test done in %lldms", t1 - t0);
LOG_VERBOSE("llm test done in %lldms", t1 - t0);
} else if (test_vit) {
// auto image = ggml_new_tensor_3d(ctx, GGML_TYPE_F32, 280, 280, 3);
// ggml_set_f32(image, 0.f);
@@ -2485,7 +2485,7 @@ namespace LLM {
// auto ref_out = load_tensor_from_file(ctx, "qwen2vl.bin");
// ggml_ext_tensor_diff(ref_out, out, 0.01f);
LOG_DEBUG("llm test done in %lldms", t1 - t0);
LOG_VERBOSE("llm test done in %lldms", t1 - t0);
} else if (test_mistral) {
std::pair<int, int> prompt_attn_range;
std::string text = "[SYSTEM_PROMPT]You are an AI that reasons about image descriptions. You give structured responses focusing on object relationships, object\nattribution and actions without speculation.[/SYSTEM_PROMPT][INST]";
@@ -2510,7 +2510,7 @@ namespace LLM {
GGML_ASSERT(!out_opt.empty());
out = std::move(out_opt);
print_sd_tensor(out);
LOG_DEBUG("llm test done in %lldms", t1 - t0);
LOG_VERBOSE("llm test done in %lldms", t1 - t0);
} else if (test_qwen3) {
std::pair<int, int> prompt_attn_range;
std::string text = "<|im_start|>user\n";
@@ -2535,7 +2535,7 @@ namespace LLM {
GGML_ASSERT(!out_opt.empty());
out = std::move(out_opt);
print_sd_tensor(out);
LOG_DEBUG("llm test done in %lldms", t1 - t0);
LOG_VERBOSE("llm test done in %lldms", t1 - t0);
} else {
std::pair<int, int> prompt_attn_range;
std::string text = "<|im_start|>system\nDescribe the image by detailing the color, shape, size, texture, quantity, text, spatial relationships of the objects and background:<|im_end|>\n<|im_start|>user\n";
@@ -2560,7 +2560,7 @@ namespace LLM {
GGML_ASSERT(!out_opt.empty());
out = std::move(out_opt);
print_sd_tensor(out);
LOG_DEBUG("llm test done in %lldms", t1 - t0);
LOG_VERBOSE("llm test done in %lldms", t1 - t0);
}
}
src/model/te/t5.hpp
+2 -2
View File
@@ -554,7 +554,7 @@ struct T5Embedder {
ss << "['" << item.first << "', " << item.second << "], ";
}
ss << "]";
LOG_DEBUG("parse '%s' to %s", text.c_str(), ss.str().c_str());
LOG_VERBOSE("parse '%s' to %s", text.c_str(), ss.str().c_str());
}
std::vector<int> tokens;
@@ -612,7 +612,7 @@ struct T5Embedder {
GGML_ASSERT(!out_opt.empty());
out = std::move(out_opt);
print_sd_tensor(out);
LOG_DEBUG("t5 test done in %lldms", t1 - t0);
LOG_VERBOSE("t5 test done in %lldms", t1 - t0);
}
}
src/model/upscaler/esrgan.hpp
+7 -7
View File
@@ -74,13 +74,13 @@ struct ESRGANConfig {
}
if (has_model_tensor || has_conv_up1 || has_conv_up2) {
LOG_DEBUG("esrgan: scale = %d, num_block = %d, num_in_ch = %d, num_out_ch = %d, num_feat = %d, num_grow_ch = %d",
config.scale,
config.num_block,
config.num_in_ch,
config.num_out_ch,
config.num_feat,
config.num_grow_ch);
LOG_VERBOSE("esrgan: scale = %d, num_block = %d, num_in_ch = %d, num_out_ch = %d, num_feat = %d, num_grow_ch = %d",
config.scale,
config.num_block,
config.num_in_ch,
config.num_out_ch,
config.num_feat,
config.num_grow_ch);
}
return config;
}
src/model/upscaler/ltx_latent_upscaler.hpp
+7 -7
View File
@@ -115,13 +115,13 @@ namespace LTXVUpsampler {
}
if (inferred) {
LOG_DEBUG("ltx latent upsampler: in_channels = %" PRId64 ", mid_channels = %" PRId64 ", num_blocks_per_stage = %d, spatial_scale = %.3f, temporal_up_factor = %d, rational_resampler = %d",
config.in_channels,
config.mid_channels,
config.num_blocks_per_stage,
config.spatial_scale,
config.temporal_up_factor,
config.rational_resampler);
LOG_VERBOSE("ltx latent upsampler: in_channels = %" PRId64 ", mid_channels = %" PRId64 ", num_blocks_per_stage = %d, spatial_scale = %.3f, temporal_up_factor = %d, rational_resampler = %d",
config.in_channels,
config.mid_channels,
config.num_blocks_per_stage,
config.spatial_scale,
config.temporal_up_factor,
config.rational_resampler);
}
return config;
}
src/model/vae/auto_encoder_kl.hpp
+2 -2
View File
@@ -864,7 +864,7 @@ struct AutoEncoderKL : public VAE {
GGML_ASSERT(!out_opt.empty());
out = std::move(out_opt);
print_sd_tensor(out);
LOG_DEBUG("encode test done in %lldms", t1 - t0);
LOG_VERBOSE("encode test done in %lldms", t1 - t0);
}
if (false) {
@@ -884,7 +884,7 @@ struct AutoEncoderKL : public VAE {
GGML_ASSERT(!out_opt.empty());
out = std::move(out_opt);
print_sd_tensor(out);
LOG_DEBUG("decode test done in %lldms", t1 - t0);
LOG_VERBOSE("decode test done in %lldms", t1 - t0);
}
};
};
src/model/vae/ltx_audio_vae.hpp
+7 -7
View File
@@ -172,12 +172,12 @@ namespace LTXV {
if (config.audio_channels != 2 || config.latent_channels != 8 || config.mel_bins != 64) {
return config;
}
LOG_DEBUG("ltx_audio_vae: sample_rate = %d, mel_bins = %d, latent_channels = %d, latent_frequency_bins = %d, has_bwe = %s",
config.sample_rate,
config.mel_bins,
config.latent_channels,
config.latent_frequency_bins,
config.has_bwe ? "true" : "false");
LOG_VERBOSE("ltx_audio_vae: sample_rate = %d, mel_bins = %d, latent_channels = %d, latent_frequency_bins = %d, has_bwe = %s",
config.sample_rate,
config.mel_bins,
config.latent_channels,
config.latent_frequency_bins,
config.has_bwe ? "true" : "false");
return config;
}
};
@@ -1063,7 +1063,7 @@ namespace LTXV {
GGML_ASSERT(!out.empty());
print_sd_tensor(out, false, "ltx_audio_vae_out");
LOG_DEBUG("ltx audio vae test done in %lldms", t1 - t0);
LOG_VERBOSE("ltx audio vae test done in %lldms", t1 - t0);
}
static void load_from_file_and_test(const std::string& model_path,
src/model/vae/ltx_vae.hpp
+17 -17
View File
@@ -1126,11 +1126,11 @@ namespace LTXVAE {
overlap, window);
overlap = window - 1;
}
LOG_DEBUG("Using temporal tiling: temporal_tile_frames = %d, temporal_tile_overlap = %d, total frames = %d, resulting in %d tiles",
window,
overlap,
(int)T,
(T + window - overlap - 1) / (window - overlap));
LOG_VERBOSE("Using temporal tiling: temporal_tile_frames = %d, temporal_tile_overlap = %d, total frames = %d, resulting in %d tiles",
window,
overlap,
(int)T,
(T + window - overlap - 1) / (window - overlap));
ggml_tensor* out = nullptr;
for (int i = 0; i < (int)T - overlap; i += (window - overlap)) {
int feat_idx = 0;
@@ -1327,21 +1327,21 @@ struct LTXVideoVAE : public VAE {
const int64_t total_frames = input.shape()[2];
auto plan = make_vae_temporal_tile_plan(total_frames, config);
LOG_DEBUG("Using streaming temporal tiling: temporal_tile_frames=%d, temporal_tile_overlap=%d, total latent frames=%lld, resulting in %d tiles",
plan.tile_frames,
plan.overlap,
(long long)total_frames,
(int)plan.tiles.size());
LOG_VERBOSE("Using streaming temporal tiling: temporal_tile_frames=%d, temporal_tile_overlap=%d, total latent frames=%lld, resulting in %d tiles",
plan.tile_frames,
plan.overlap,
(long long)total_frames,
(int)plan.tiles.size());
free_cache_ctx_and_buffer();
auto output = process_vae_temporal_tiles(input, plan, [&](const sd::Tensor<float>& z_chunk, const VAETemporalTile& tile) {
LOG_DEBUG("LTX VAE temporal tile %lld/%d: latent frames [%lld, %lld), overlap=%d",
(long long)tile.index + 1,
(int)plan.tiles.size(),
(long long)tile.start,
(long long)tile.end,
tile.overlap);
LOG_VERBOSE("LTX VAE temporal tile %lld/%d: latent frames [%lld, %lld), overlap=%d",
(long long)tile.index + 1,
(int)plan.tiles.size(),
(long long)tile.start,
(long long)tile.end,
tile.overlap);
auto get_graph = [&]() -> ggml_cgraph* {
return build_temporal_tile_graph(z_chunk,
@@ -1465,7 +1465,7 @@ struct LTXVideoVAE : public VAE {
GGML_ASSERT(!out.empty());
print_sd_tensor(out, false, "ltx_vae_out");
LOG_DEBUG("ltx vae test done in %lldms", t1 - t0);
LOG_VERBOSE("ltx vae test done in %lldms", t1 - t0);
}
static void load_from_file_and_test(const std::string& model_path,
src/model/vae/tae.hpp
+1 -1
View File
@@ -65,7 +65,7 @@ public:
if (n_in != n_out) {
auto skip = std::dynamic_pointer_cast<Conv2d>(blocks["skip"]);
LOG_DEBUG("skip");
LOG_VERBOSE("skip");
x = skip->forward(ctx, x);
}
src/model/vae/vae.hpp
+16 -16
View File
@@ -54,23 +54,23 @@ protected:
}
auto plan = make_vae_temporal_tile_plan(input.shape()[2], resolved_config);
LOG_DEBUG("%s temporal tiling: tile_frames=%d, overlap=%d, total_frames=%lld, tiles=%d",
get_desc().c_str(),
plan.tile_frames,
plan.overlap,
(long long)input.shape()[2],
(int)plan.tiles.size());
LOG_VERBOSE("%s temporal tiling: tile_frames=%d, overlap=%d, total_frames=%lld, tiles=%d",
get_desc().c_str(),
plan.tile_frames,
plan.overlap,
(long long)input.shape()[2],
(int)plan.tiles.size());
return process_vae_temporal_tiles_blended(
input,
plan,
output_scale,
[&](const sd::Tensor<float>& input_tile, const VAETemporalTile& tile) {
LOG_DEBUG("%s temporal tile %d/%d: input frames [%lld, %lld)",
get_desc().c_str(),
tile.index + 1,
(int)plan.tiles.size(),
(long long)tile.start,
(long long)tile.end);
LOG_VERBOSE("%s temporal tile %d/%d: input frames [%lld, %lld)",
get_desc().c_str(),
tile.index + 1,
(int)plan.tiles.size(),
(long long)tile.start,
(long long)tile.end);
return _compute(n_threads, input_tile, true);
});
}
@@ -230,7 +230,7 @@ public:
const float encode_tile_factor = sd_version_is_minimax_h3(version) ? 1.f : (sd_version_is_wan(version) || sd_version_is_hunyuan_video(version) || sd_version_is_ltxav(version)) ? 1.30539f
: 2.0f;
get_tile_sizes(tile_size_x, tile_size_y, tile_overlap, tiling_params, W, H, encode_tile_factor);
LOG_DEBUG("VAE Tile size: %dx%d", tile_size_x, tile_size_y);
LOG_VERBOSE("VAE Tile size: %dx%d", tile_size_x, tile_size_y);
output = tiled_compute(input,
n_threads,
static_cast<int>(W),
@@ -258,7 +258,7 @@ public:
return {};
}
int64_t t1 = ggml_time_ms();
LOG_DEBUG("computing vae encode graph completed, taking %.2fs", (t1 - t0) * 1.0f / 1000);
LOG_VERBOSE("computing vae encode graph completed, taking %.2fs", (t1 - t0) * 1.0f / 1000);
return std::move(output);
}
@@ -281,7 +281,7 @@ public:
int tile_size_x, tile_size_y;
get_tile_sizes(tile_size_x, tile_size_y, tile_overlap, tiling_params, input.shape()[0], input.shape()[1]);
if (!silent) {
LOG_DEBUG("VAE Tile size: %dx%d", tile_size_x, tile_size_y);
LOG_VERBOSE("VAE Tile size: %dx%d", tile_size_x, tile_size_y);
}
output = tiled_compute(
input,
@@ -315,7 +315,7 @@ public:
scale_tensor_to_0_1(&output);
}
int64_t t1 = ggml_time_ms();
LOG_DEBUG("computing vae decode graph completed, taking %.2fs", (t1 - t0) * 1.0f / 1000);
LOG_VERBOSE("computing vae decode graph completed, taking %.2fs", (t1 - t0) * 1.0f / 1000);
return std::move(output);
}
src/model/vae/wan_vae.hpp
+11 -11
View File
@@ -1278,7 +1278,7 @@ namespace WAN {
}
}
if (is_2D) {
LOG_DEBUG("USING 2D VAE");
LOG_VERBOSE("USING 2D VAE");
}
ae = WanVAE(decode_only, version, is_2D);
ae.init(params_ctx, tensor_storage_map, prefix);
@@ -1409,20 +1409,20 @@ namespace WAN {
stateful_config.overlap = 0;
auto plan = make_vae_temporal_tile_plan(input.shape()[2], stateful_config);
LOG_DEBUG("Wan VAE stateful temporal tiling: tile_frames=%d, total latent frames=%lld, tiles=%d",
plan.tile_frames,
(long long)input.shape()[2],
(int)plan.tiles.size());
LOG_VERBOSE("Wan VAE stateful temporal tiling: tile_frames=%d, total latent frames=%lld, tiles=%d",
plan.tile_frames,
(long long)input.shape()[2],
(int)plan.tiles.size());
free_cache_ctx_and_buffer();
ae.clear_cache();
auto output = process_vae_temporal_tiles(input, plan, [&](const sd::Tensor<float>& input_tile, const VAETemporalTile& tile) {
LOG_DEBUG("Wan VAE temporal tile %d/%d: latent frames [%lld, %lld)",
tile.index + 1,
(int)plan.tiles.size(),
(long long)tile.start,
(long long)tile.end);
LOG_VERBOSE("Wan VAE temporal tile %d/%d: latent frames [%lld, %lld)",
tile.index + 1,
(int)plan.tiles.size(),
(long long)tile.start,
(long long)tile.end);
auto get_graph = [&]() -> ggml_cgraph* {
return build_temporal_tile_graph(input_tile, static_cast<int>(tile.start));
};
@@ -1479,7 +1479,7 @@ namespace WAN {
GGML_ASSERT(!out_opt.empty());
out = std::move(out_opt);
print_sd_tensor(out);
LOG_DEBUG("decode test done in %ldms", t1 - t0);
LOG_VERBOSE("decode test done in %ldms", t1 - t0);
}
};