ComfyUI Node
ASV Prompt Generator
A ComfyUI node in Prompt with 8 inputs and 1 output.
ASV Prompt Generator
- STRING
◄seed825628174325066►
◄enabledtrue►
◄prompt►
◄samplingtrue►
◄temperature1.20►
◄top_k50►
◄top_p0.90►
◄notesREADME:
* Sampling: Enables random sampling, allowing the model to generate more varied and creative outputs by choosing from a range of probable tokens instead of only the most likely one.
* Temperature: Controls the randomness of predictions by scaling the logits before applying softmax. Higher values (e.g., 0.8 to 1.5) result in more randomness, while values closer to 0 make the output more deterministic.
* Top_k: Limits the sample pool to the top k highest-probability next tokens, adding randomness by only sampling from this subset.
* Top_p: Implements 'nucleus sampling', which includes tokens with a cumulative probability above a threshold (e.g., 0.9), allowing for dynamic vocabulary selection based on the context.
Important: This textbox serves as a README and a personal note-taking space. It is solely for reference and will not impact image generation in any way.
https://github.com/zubenelakrab/ComfyUI-ASV-Nodes►
CategoryPrompt
Inputs (8)
| Name | Type | Default | Description |
|---|---|---|---|
| seed | INT | 8256281743250660–999999999999999 | — |
| enabled | BOOLEAN | true | — |
| prompt | STRING | — | |
| sampling | BOOLEAN | true | — |
| temperature | FLOAT | 1.200–1.5 | — |
| top_k | INT | 501–100 | — |
| top_p | FLOAT | 0.900–1 | — |
| notes | STRING | README: * Sampling: Enables random sampling, allowing the model to generate more varied and creative outputs by choosing from a range of probable tokens instead of only the most likely one. * Temperature: Controls the randomness of predictions by scaling the logits before applying softmax. Higher values (e.g., 0.8 to 1.5) result in more randomness, while values closer to 0 make the output more deterministic. * Top_k: Limits the sample pool to the top k highest-probability next tokens, adding randomness by only sampling from this subset. * Top_p: Implements 'nucleus sampling', which includes tokens with a cumulative probability above a threshold (e.g., 0.9), allowing for dynamic vocabulary selection based on the context. Important: This textbox serves as a README and a personal note-taking space. It is solely for reference and will not impact image generation in any way. https://github.com/zubenelakrab/ComfyUI-ASV-Nodes | — |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| STRING | STRING | — |