Instructions to use saik0s/comfy_backup with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- llama-cpp-python
How to use saik0s/comfy_backup with llama-cpp-python:
# !pip install llama-cpp-python from llama_cpp import Llama llm = Llama.from_pretrained( repo_id="saik0s/comfy_backup", filename="ComfyUI/models/text_encoders/gemma-3-12b-it-q2_k.gguf", )
llm.create_chat_completion( messages = "No input example has been defined for this model task." )
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- llama.cpp
How to use saik0s/comfy_backup with llama.cpp:
Install (macOS, Linux)
curl -LsSf https://llama.app/install.sh | sh # Start a local OpenAI-compatible server with a web UI: llama serve -hf saik0s/comfy_backup:Q4_K_S # Run inference directly in the terminal: llama cli -hf saik0s/comfy_backup:Q4_K_S
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf saik0s/comfy_backup:Q4_K_S # Run inference directly in the terminal: llama cli -hf saik0s/comfy_backup:Q4_K_S
Use pre-built binary
# Download pre-built binary from: # https://github.com/ggerganov/llama.cpp/releases # Start a local OpenAI-compatible server with a web UI: ./llama-server -hf saik0s/comfy_backup:Q4_K_S # Run inference directly in the terminal: ./llama-cli -hf saik0s/comfy_backup:Q4_K_S
Build from source code
git clone https://github.com/ggerganov/llama.cpp.git cd llama.cpp cmake -B build cmake --build build -j --target llama-server llama-cli # Start a local OpenAI-compatible server with a web UI: ./build/bin/llama-server -hf saik0s/comfy_backup:Q4_K_S # Run inference directly in the terminal: ./build/bin/llama-cli -hf saik0s/comfy_backup:Q4_K_S
Use Docker
docker model run hf.co/saik0s/comfy_backup:Q4_K_S
- LM Studio
- Jan
- Ollama
How to use saik0s/comfy_backup with Ollama:
ollama run hf.co/saik0s/comfy_backup:Q4_K_S
- Unsloth Studio
How to use saik0s/comfy_backup with Unsloth Studio:
Install Unsloth Studio (macOS, Linux, WSL)
curl -fsSL https://unsloth.ai/install.sh | sh # Run unsloth studio unsloth studio -H 0.0.0.0 -p 8888 # Then open http://localhost:8888 in your browser # Search for saik0s/comfy_backup to start chatting
Install Unsloth Studio (Windows)
irm https://unsloth.ai/install.ps1 | iex # Run unsloth studio unsloth studio -H 0.0.0.0 -p 8888 # Then open http://localhost:8888 in your browser # Search for saik0s/comfy_backup to start chatting
Using HuggingFace Spaces for Unsloth
# No setup required # Open https://huggingface.co/spaces/unsloth/studio in your browser # Search for saik0s/comfy_backup to start chatting
- Pi
How to use saik0s/comfy_backup with Pi:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf saik0s/comfy_backup:Q4_K_S
Configure the model in Pi
# Install Pi: npm install -g @mariozechner/pi-coding-agent # Add to ~/.pi/agent/models.json: { "providers": { "llama-cpp": { "baseUrl": "http://localhost:8080/v1", "api": "openai-completions", "apiKey": "none", "models": [ { "id": "saik0s/comfy_backup:Q4_K_S" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Hermes Agent new
How to use saik0s/comfy_backup with Hermes Agent:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf saik0s/comfy_backup:Q4_K_S
Configure Hermes
# Install Hermes: curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash hermes setup # Point Hermes at the local server: hermes config set model.provider custom hermes config set model.base_url http://127.0.0.1:8080/v1 hermes config set model.default saik0s/comfy_backup:Q4_K_S
Run Hermes
hermes
- Atomic Chat new
- OpenClaw new
How to use saik0s/comfy_backup with OpenClaw:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf saik0s/comfy_backup:Q4_K_S
Configure OpenClaw
# Install OpenClaw: npm install -g openclaw@latest # Register the local server and set it as the default model: openclaw onboard --non-interactive --mode local \ --auth-choice custom-api-key \ --custom-base-url http://127.0.0.1:8080/v1 \ --custom-model-id "saik0s/comfy_backup:Q4_K_S" \ --custom-provider-id llama-cpp \ --custom-compatibility openai \ --custom-text-input \ --accept-risk \ --skip-health
Run OpenClaw
openclaw agent --local --agent main --message "Hello from Hugging Face"
- Docker Model Runner
How to use saik0s/comfy_backup with Docker Model Runner:
docker model run hf.co/saik0s/comfy_backup:Q4_K_S
- Lemonade
How to use saik0s/comfy_backup with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull saik0s/comfy_backup:Q4_K_S
Run and chat with the model
lemonade run user.comfy_backup-Q4_K_S
List all available models
lemonade list
| import type { | |
| LGraphNode, | |
| LLink, | |
| LGraphCanvas, | |
| INodeInputSlot, | |
| INodeOutputSlot, | |
| ISlotType, | |
| } from "@comfyorg/frontend"; | |
| import type {ComfyNodeDef} from "typings/comfy.js"; | |
| import {app} from "scripts/app.js"; | |
| import {DynamicContextNodeBase, InputLike} from "./dynamic_context_base.js"; | |
| import {NodeTypesString} from "./constants.js"; | |
| import { | |
| InputMutation, | |
| SERVICE as CONTEXT_SERVICE, | |
| getContextOutputName, | |
| } from "./services/context_service.js"; | |
| import {getConnectedInputNodesAndFilterPassThroughs} from "./utils.js"; | |
| import {debounce, moveArrayItem} from "rgthree/common/shared_utils.js"; | |
| import {measureText} from "./utils_canvas.js"; | |
| import {SERVICE as CONFIG_SERVICE} from "./services/config_service.js"; | |
| type ShadowInputData = { | |
| node: LGraphNode; | |
| slot: number; | |
| shadowIndex: number; | |
| shadowIndexIfShownSingularly: number; | |
| shadowIndexFull: number; | |
| nodeIndex: number; | |
| type: string | -1; | |
| name: string; | |
| key: string; | |
| // isDuplicatedBefore: boolean, | |
| duplicatesBefore: number[]; | |
| duplicatesAfter: number[]; | |
| }; | |
| /** | |
| * The Context Switch node. | |
| */ | |
| class DynamicContextSwitchNode extends DynamicContextNodeBase { | |
| static override title = NodeTypesString.DYNAMIC_CONTEXT_SWITCH; | |
| static override type = NodeTypesString.DYNAMIC_CONTEXT_SWITCH; | |
| static comfyClass = NodeTypesString.DYNAMIC_CONTEXT_SWITCH; | |
| protected override readonly hasShadowInputs = true; | |
| // override hasShadowInputs = true; | |
| /** | |
| * We should be able to assume that `lastInputsList` is the input list after the last, major | |
| * synchronous change. Which should mean, if we're handling a change that is currently live, but | |
| * not represented in our node (like, an upstream node has already removed an input), then we | |
| * should be able to compar the current InputList to this `lastInputsList`. | |
| */ | |
| lastInputsList: ShadowInputData[] = []; | |
| private shadowInputs: (InputLike & {count: number})[] = [ | |
| {name: "base_ctx", type: "RGTHREE_DYNAMIC_CONTEXT", link: null, count: 0, boundingRect: null}, | |
| ]; | |
| constructor(title = DynamicContextSwitchNode.title) { | |
| super(title); | |
| } | |
| override getContextInputsList() { | |
| return this.shadowInputs; | |
| } | |
| override handleUpstreamMutation(mutation: InputMutation) { | |
| this.scheduleHardRefresh(); | |
| } | |
| override onConnectionsChange( | |
| type: ISlotType, | |
| slotIndex: number, | |
| isConnected: boolean, | |
| link: LLink | null | undefined, | |
| inputOrOutput: INodeInputSlot | INodeOutputSlot, | |
| ): void { | |
| super.onConnectionsChange?.call(this, type, slotIndex, isConnected, link, inputOrOutput); | |
| if (this.configuring) { | |
| return; | |
| } | |
| if (type === LiteGraph.INPUT) { | |
| this.scheduleHardRefresh(); | |
| } | |
| } | |
| scheduleHardRefresh(ms = 64) { | |
| return debounce(() => { | |
| this.refreshInputsAndOutputs(); | |
| }, ms); | |
| } | |
| override onNodeCreated() { | |
| this.addInput("ctx_1", "RGTHREE_DYNAMIC_CONTEXT"); | |
| this.addInput("ctx_2", "RGTHREE_DYNAMIC_CONTEXT"); | |
| this.addInput("ctx_3", "RGTHREE_DYNAMIC_CONTEXT"); | |
| this.addInput("ctx_4", "RGTHREE_DYNAMIC_CONTEXT"); | |
| this.addInput("ctx_5", "RGTHREE_DYNAMIC_CONTEXT"); | |
| super.onNodeCreated(); | |
| } | |
| override addContextInput(name: string, type: string, slot?: number): void {} | |
| /** | |
| * This is a "hard" refresh of the list, but looping over the actual context inputs, and | |
| * recompiling the shadowInputs and outputs. | |
| */ | |
| private refreshInputsAndOutputs() { | |
| const inputs: (InputLike & {count: number})[] = [ | |
| {name: "base_ctx", type: "RGTHREE_DYNAMIC_CONTEXT", link: null, count: 0, boundingRect: null}, | |
| ]; | |
| let numConnected = 0; | |
| for (let i = 0; i < this.inputs.length; i++) { | |
| const childCtxs = getConnectedInputNodesAndFilterPassThroughs( | |
| this, | |
| this, | |
| i, | |
| ) as DynamicContextNodeBase[]; | |
| if (childCtxs.length > 1) { | |
| throw new Error("How is there more than one input?"); | |
| } | |
| const ctx = childCtxs[0]; | |
| if (!ctx) continue; | |
| numConnected++; | |
| const slotsData = CONTEXT_SERVICE.getDynamicContextInputsData(ctx); | |
| console.log(slotsData); | |
| for (const slotData of slotsData) { | |
| const found = inputs.find( | |
| (n) => getContextOutputName(slotData.name) === getContextOutputName(n.name), | |
| ); | |
| if (found) { | |
| found.count += 1; | |
| continue; | |
| } | |
| inputs.push({ | |
| name: slotData.name, | |
| type: slotData.type, | |
| link: null, | |
| count: 1, | |
| boundingRect: null, | |
| }); | |
| } | |
| } | |
| this.shadowInputs = inputs; | |
| // First output is always CONTEXT, so "p" is the offset. | |
| let i = 0; | |
| for (i; i < this.shadowInputs.length; i++) { | |
| const data = this.shadowInputs[i]!; | |
| let existing = this.outputs.find( | |
| (o) => getContextOutputName(o.name) === getContextOutputName(data.name), | |
| ); | |
| if (!existing) { | |
| existing = this.addOutput(getContextOutputName(data.name), data.type); | |
| } | |
| moveArrayItem(this.outputs, existing, i); | |
| delete existing.rgthree_status; | |
| if (data.count !== numConnected) { | |
| existing.rgthree_status = "WARN"; | |
| } | |
| } | |
| while (this.outputs[i]) { | |
| const output = this.outputs[i]; | |
| if (output?.links?.length) { | |
| output.rgthree_status = "ERROR"; | |
| i++; | |
| } else { | |
| this.removeOutput(i); | |
| } | |
| } | |
| this.fixInputsOutputsLinkSlots(); | |
| } | |
| override onDrawForeground(ctx: CanvasRenderingContext2D, canvas: LGraphCanvas): void { | |
| const low_quality = (canvas?.ds?.scale ?? 1) < 0.6; | |
| if (low_quality || this.size[0] <= 10) { | |
| return; | |
| } | |
| let y = LiteGraph.NODE_SLOT_HEIGHT - 1; | |
| const w = this.size[0]; | |
| ctx.save(); | |
| ctx.font = "normal " + LiteGraph.NODE_SUBTEXT_SIZE + "px Arial"; | |
| ctx.textAlign = "right"; | |
| for (const output of this.outputs) { | |
| if (!output.rgthree_status) { | |
| y += LiteGraph.NODE_SLOT_HEIGHT; | |
| continue; | |
| } | |
| const x = w - 20 - measureText(ctx, output.name); | |
| if (output.rgthree_status === "ERROR") { | |
| ctx.fillText("🛑", x, y); | |
| } else if (output.rgthree_status === "WARN") { | |
| ctx.fillText("⚠️", x, y); | |
| } | |
| y += LiteGraph.NODE_SLOT_HEIGHT; | |
| } | |
| ctx.restore(); | |
| } | |
| } | |
| app.registerExtension({ | |
| name: "rgthree.DynamicContextSwitch", | |
| async beforeRegisterNodeDef(nodeType: typeof LGraphNode, nodeData: ComfyNodeDef) { | |
| if (!CONFIG_SERVICE.getConfigValue("unreleased.dynamic_context.enabled")) { | |
| return; | |
| } | |
| if (nodeData.name === DynamicContextSwitchNode.type) { | |
| DynamicContextSwitchNode.setUp(nodeType, nodeData); | |
| } | |
| }, | |
| }); | |