Load LoRA from file path

This does exactly what ComfyUI's built-in LoraLoader does - patch a diffusion model and its CLIP encoder with a LoRA so it can render a specific style, character, or concept - except instead of picking from a dropdown that only sees files inside ComfyUI's own models/loras folder, you give it a raw path and filename. That's the entire pitch: if your LoRAs live somewhere ComfyUI doesn't automatically scan - a shared network drive, a folder you're managing outside the standard layout, or a path you want to drive from a variable elsewhere in a bigger pipeline - this node reaches it without you having to symlink or copy anything into place first.

A LoRA is a small patch, not a full model: it freezes the base checkpoint and adds low-rank adjustments on top, which is why file sizes run tens to a couple hundred megabytes instead of gigabytes. This node needs both halves of what a LoRA touches, which is why model and clip are separate required inputs - LoRAs commonly modify the diffusion model and the CLIP text encoder differently, so you get independent strength controls for each: strength_model and strength_clip, both defaulting to 1 and both able to go negative if you actually want to subtract a LoRA's effect rather than add it. lora_path and lora_name are split into two fields rather than one combined path - lora_path is the directory (defaulting to a Windows-style D:/LORA, a tell that this pack's primary dev environment is Windows) and lora_name is the filename within it (defaulting to lora.pt). Outputs are the modified MODEL and CLIP, which chain onward exactly like the stock loader's outputs do - into your sampler and text encoder respectively.

One thing worth flagging on the default: lora.pt as a placeholder filename is a .pt extension, but the overwhelming majority of LoRAs you'll actually download - from CivitAI or anywhere else - ship as .safetensors. That's not a limitation of this node, just don't assume the default extension is telling you what format to expect; point lora_name at whatever your actual file is named, .safetensors included.

Installing it means installing the whole pack, since this node ships as part of comfyui_LLM_party: search "comfyui_LLM_party" in ComfyUI Manager, or git clone https://github.com/heshengtao/comfyui_LLM_party into custom_nodes and restart, then run pip install -r requirements.txt from inside the pack folder using ComfyUI's own Python (portable installs: python_embeded\python.exe -m pip install -r requirements.txt).

If you're just using local LoRAs from the standard models/loras folder, the core LoraLoader node does the same job with a friendlier dropdown and no risk of a typo'd path - reach for this one specifically when your files genuinely live somewhere else. And remember the general LoRA rule that trips people up regardless of which loader you use: LoRAs are architecture-bound, so an SDXL LoRA does nothing useful on a Flux checkpoint and vice versa - if the output looks like the LoRA simply isn't applying at all, check that it actually matches your base model's architecture before you start second-guessing the strength values.