ComfyUI Node
IF Image to Prompt🖼️
A ComfyUI node in ImpactFrames💥🎞️/IF_tools with 32 inputs and 6 outputs.
IF Image to Prompt🖼️
- images
- mask
- Omni
- question
- response
- negative
- omni
- generated_images
- mask
◄llm_provider▾►
◄llm_model▾►
◄base_iplocalhost►
◄port11434►
◄user_prompt—►
◄strategynormal►
◄prime_directives—►
◄profilesNone►
◄embellish_prompt▾►
◄style_prompt▾►
◄neg_prompt▾►
◄stop_string▾►
◄max_tokens2048►
◄randomfalse►
◄seed0►
◄temperature0.70►
◄top_k40►
◄top_p0.90►
◄repeat_penalty1.10►
◄keep_alivefalse►
◄clear_historyfalse►
◄history_steps10►
◄aspect_ratio1:1►
◄autofalse►
◄auto_modefalse►
◄batch_count4►
◄external_api_key►
◄precision▾►
◄attention▾►
CategoryImpactFrames💥🎞️/IF_tools
Inputs (32)
| Name | Type | Default | Description |
|---|---|---|---|
| images | IMAGE | — | |
| llm_provider | COMBO | 12 options: xai, llamacpp, ollama, kobold, lmstudio, textgen, +6 | |
| llm_model | COMBO | 0 options: | |
| base_ip | STRING | localhost | — |
| port | STRING | 11434 | — |
| user_prompt | STRING | — | |
| strategyopt | COMBO | normal | 5 options: normal, omost, create, edit, variations |
| maskopt | MASK | — | |
| prime_directivesopt | STRING | The system prompt for the LLM. | |
| profilesopt | COMBO | None | The pre-defined system_prompt from the json profile file on the presets folder you can edit or make your own will be listed here. |
| embellish_promptopt | COMBO | The pre-defined embellishment from the json embellishments file on the presets folder you can edit or make your own will be listed here. | |
| style_promptopt | COMBO | The pre-defined style from the json style_prompts file on the presets folder you can edit or make your own will be listed here. | |
| neg_promptopt | COMBO | The pre-defined negative prompt from the json neg_prompts file on the presets folder you can edit or make your own will be listed here. | |
| stop_stringopt | COMBO | Specifies a string at which text generation should stop. | |
| max_tokensopt | INT | 20481–8192 | Maximum number of tokens to generate in the response. |
| randomopt | BOOLEAN | false | Toggles between using a fixed seed or temperature-based randomness. |
| seedopt | INT | 0 | Random seed for reproducible outputs. |
| temperatureopt | FLOAT | 0.700–1 | Controls randomness in output generation. Higher values increase creativity but may reduce coherence. |
| top_kopt | INT | 40 | Limits the next token selection to the K most likely tokens. |
| top_popt | FLOAT | 0.90 | Cumulative probability cutoff for token selection. |
| repeat_penaltyopt | FLOAT | 1.10 | Penalizes repetition in generated text. |
| keep_aliveopt | BOOLEAN | false | Determines whether to keep the model loaded in memory between calls. |
| clear_historyopt | BOOLEAN | false | Determines whether to clear the history between calls. |
| history_stepsopt | INT | 10 | Number of steps to keep in history. |
| aspect_ratioopt | COMBO | 1:1 | Aspect ratio for the generated images. |
| autoopt | BOOLEAN | false | If true, it generates auto promts based on the listed images click the save combomix settings to set the auto prompt generation file |
| auto_modeopt | BOOLEAN | false | If true, it generates a prompt for each image with Combo mode and Mix mode combined a maximum of 4 images in the list then moves to the next 4 and use it to run a job as many times as your batch count is set. the settings are taken from the yaml file |
| batch_countopt | INT | 4 | Number of images to generate. only for create, edit and variations strategies. |
| external_api_keyopt | STRING | If this is not empty, it will be used instead of the API key from the .env file. Make sure it is empty to use the .env file. | |
| precisionopt | COMBO | Select preccision on Transformer models. | |
| attentionopt | COMBO | Select attention mechanism on Transformer models. | |
| Omniopt | OMNI | Additional input for the selected tool. |
Outputs (6)
| Name | Type | Description |
|---|---|---|
| question | STRING | — |
| response | STRING | — |
| negative | STRING | — |
| omni | OMNI | — |
| generated_images | IMAGE | — |
| mask | MASK | — |