diff --git "a/yy_krea2_lokr_v1/kontol.log" "b/yy_krea2_lokr_v1/kontol.log" new file mode 100644--- /dev/null +++ "b/yy_krea2_lokr_v1/kontol.log" @@ -0,0 +1,735 @@ +Running 1 job +{ + "type": "diffusion_trainer", + "training_folder": "/root/output_lora_ostris", + "device": "cuda", + "trigger_word": "ohwx", + "performance_log_every": 500, + "network": { + "type": "lokr", + "linear": 32, + "linear_alpha": 32, + "conv": 16, + "conv_alpha": 16, + "lokr_full_rank": true, + "lokr_factor": 8, + "network_kwargs": { + "ignore_if_contains": [] + } + }, + "save": { + "dtype": "fp32", + "save_every": 800, + "max_step_saves_to_keep": 30, + "save_format": "diffusers", + "push_to_hub": false + }, + "datasets": [ + { + "folder_path": "/root/dataset-24", + "mask_path": null, + "mask_min_value": 0.1, + "default_caption": "", + "caption_ext": "txt", + "caption_dropout_rate": 0.05, + "cache_latents_to_disk": true, + "is_reg": false, + "network_weight": 1, + "resolution": [ + 512, + 768, + 1024 + ], + "controls": [], + "shrink_video_to_frames": true, + "num_frames": 1, + "flip_x": false, + "flip_y": false, + "num_repeats": 1 + } + ], + "train": { + "batch_size": 1, + "bypass_guidance_embedding": false, + "steps": 3000, + "gradient_accumulation": 1, + "train_unet": true, + "train_text_encoder": false, + "gradient_checkpointing": false, + "noise_scheduler": "flowmatch", + "optimizer": "automagic2", + "timestep_type": "sigmoid", + "content_or_style": "balanced", + "optimizer_params": { + "weight_decay": 0.0001 + }, + "unload_text_encoder": true, + "cache_text_embeddings": false, + "lr": 1e-06, + "ema_config": { + "use_ema": false, + "ema_decay": 0.99 + }, + "skip_first_sample": true, + "force_first_sample": false, + "disable_sampling": false, + "dtype": "bf16", + "diff_output_preservation": false, + "diff_output_preservation_multiplier": 1, + "diff_output_preservation_class": "person", + "switch_boundary_every": 1, + "loss_type": "mse", + "do_differential_guidance": true, + "differential_guidance_scale": 3 + }, + "logging": { + "log_every": 1, + "use_ui_logger": false + }, + "model": { + "name_or_path": "/root/checkpoint", + "quantize": false, + "qtype": "qfloat8", + "quantize_te": false, + "qtype_te": "qfloat8", + "arch": "krea2", + "low_vram": true, + "model_kwargs": {}, + "compile": true, + "layer_offloading": false, + "layer_offloading_text_encoder_percent": 1, + "layer_offloading_transformer_percent": 1 + }, + "sample": { + "sampler": "flowmatch", + "sample_every": 1600, + "width": 1024, + "height": 1024, + "samples": [ + { + "prompt": "amateur (photos taken in 2010s), medium-shot, wearing nothing on torso, sweaty ohwx woman dancing energetically at a concert in the noon. , She is wearing no makeup" + }, + { + "prompt": "an overhead close-up photo of a face of ohwx woman in a seedy toilet. She is nude, covering her bare tits using her hands, looking at camera above sultrily while sticking out tongue. 35mm film grain, Terry Richardson-style photography, erotic vibes." + }, + { + "prompt": "a high-quality, medium-shot, maternity, wearing nothing on torso, boudoir photoshoot of ohwx woman in nature. She is curvy. She is looking at the camera." + }, + { + "prompt": "an amateur, shaky, grainy and noisy, waist-up medium shot inside a sauna of ohwx with messy hair, frontal perspective. She is slim and completely naked, grabbing strongly the wood chair in front of her. A masked man from behind hugs the woman, puts his hands over the woman's chest, squeezing the woman's tits from behind roughly as he kisses the woman's neck passionately. THe woman is moaning in pleasure while looking at the camera" + } + ], + "neg": "", + "seed": 42525272726, + "walk_seed": false, + "guidance_scale": 1, + "sample_steps": 8, + "num_frames": 1, + "fps": 1 + } +} + +############################################# +# Running job: yy_krea2_lokr_v1 +############################################# + + +Running 1 process +Loading Krea 2 model +Loading transformer (SingleStreamDiT) + - fetching transformer weights + - loading transformer state dict +Moving transformer to CPU +Loading Qwen3-VL text encoder from Qwen/Qwen3-VL-4B-Instruct + config.json: 0.00B [00:00, ?B/s] config.json: 0.00B [00:00, ?B/s] config.json: 1.50kB [00:00, 3.60MB/s] config.json: 1.50kB [00:00, 3.60MB/s] + + tokenizer_config.json: 0.00B [00:00, ?B/s] tokenizer_config.json: 0.00B [00:00, ?B/s] tokenizer_config.json: 10.9kB [00:00, 26.6MB/s] tokenizer_config.json: 10.9kB [00:00, 26.6MB/s] + + vocab.json: 0.00B [00:00, ?B/s] vocab.json: 0.00B [00:00, ?B/s] vocab.json: 2.78MB [00:00, 21.3MB/s] vocab.json: 2.78MB [00:00, 21.3MB/s] vocab.json: 2.78MB [00:00, 20.9MB/s] vocab.json: 2.78MB [00:00, 20.9MB/s] + + merges.txt: 0.00B [00:00, ?B/s] merges.txt: 0.00B [00:00, ?B/s] merges.txt: 1.67MB [00:00, 19.8MB/s] merges.txt: 1.67MB [00:00, 19.8MB/s] + + tokenizer.json: 0.00B [00:00, ?B/s] tokenizer.json: 0.00B [00:00, ?B/s] tokenizer.json: 7.03MB [00:00, 31.0MB/s] tokenizer.json: 7.03MB [00:00, 31.0MB/s] tokenizer.json: 7.03MB [00:00, 30.5MB/s] tokenizer.json: 7.03MB [00:00, 30.5MB/s] + + model.safetensors.index.json: 0.00B [00:00, ?B/s] model.safetensors.index.json: 0.00B [00:00, ?B/s] model.safetensors.index.json: 64.7kB [00:00, 135MB/s] model.safetensors.index.json: 64.7kB [00:00, 135MB/s] + + Downloading (incomplete total...): 0.00B [00:00, ?B/s] Downloading (incomplete total...): 0.00B [00:00, ?B/s] + + Fetching 2 files: 0%| | 0/2 [00:00