<ul><li>Trigger Word: <code>sks_rosetta</code></li><li>LoRA Weight: <code>1.0</code> (Recommended)</li><li>Distilled LoRA Weight: <code>0.25</code></li><li>Sampling Steps: <code>25</code></li><li>CFG Scale: <code>1.5</code></li><li>Training Resolution: <code>1024x1024</code></li></ul>datasets di kumpulkan dari chatgptyang di civitai versi steps 2500 bs1coba steps lainya menggunakan bs 2 di HF <a target="_blank" rel="ugc" href="https://huggingface.co/Heouzen/LTX-2.3_sks_rosetta_lora_r32">rosetta</a>

LTX-2.3_sks_rosetta_indonesia

<pre><code> trigger_word: "sks_rosetta"
 network:
 type: "lora"
 linear: 32
 linear_alpha: 32
 conv: 16
 conv_alpha: 16
 lokr_full_rank: true
 lokr_factor: -1
 datasets:
 resolution:
 - 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: true
 noise_scheduler: "flowmatch"
 optimizer: "adamw"
 timestep_type: "linear"
 content_or_style: "balanced"
 optimizer_params:
 weight_decay: 0.0001
 unload_text_encoder: false
 cache_text_embeddings: true
 lr: 0.0001
 ema_config:
 use_ema: false
 ema_decay: 0.99
 dtype: "bf16"
</code></pre>

Files with SHA256: 8585c1388e6b7ed1687ae6c2b0f4c40401c06d806cb43270f6003517be2f5e2e