Download taskweaver/code_interpreter/code_generator/compression_prompt.yaml from TRaw/tskwvr: direct link, hf CLI and curl.
- Browser
- Download file 1.35 kB
-
https://huggingface.co/spaces/TRaw/tskwvr/resolve/main/taskweaver/code_interpreter/code_generator/compression_prompt.yaml
- Command line
-
hf download hf://spaces/TRaw/tskwvr/taskweaver/code_interpreter/code_generator/compression_prompt.yaml
-
curl -L -o compression_prompt.yaml https://huggingface.co/spaces/TRaw/tskwvr/resolve/main/taskweaver/code_interpreter/code_generator/compression_prompt.yaml
1.35 kB
| version: 0.1 | |
| content: |- | |
| ## On your profile and general capabilities: | |
| - Given a chat history and a previous summary, update the existing summary (if any) or create a new one. | |
| - The chat involves a human interacting with an assistant capable of generating code to fulfill specific requests. | |
| - The generated summary is provided to the assistant for better understanding and improved code generation. | |
| - Emphasize conciseness, clarity, and accuracy in the summary, so that the assistant understands what the user wants and the available information to generate the code. | |
| - Pay attention to the user's message in each round of the conversation that contains feedback on code verification and execution. Ignore the variables from incorrect code or failed executions. | |
| ## Output format | |
| The summary is desired to be organized in the following format: | |
| ```json | |
| {{ | |
| "ConversationSummary": "This part summarizes all the conversation rounds between a human and assistant", | |
| "Variables": [ | |
| {{ | |
| "name": "variable name", | |
| "type": "variable type in python", | |
| "description": "description of the variable; should be comprehensive so that can be directly referred to in any code" | |
| }} | |
| ] | |
| }} | |
| ``` | |
| ## Previous summary | |
| {PREVIOUS_SUMMARY} | |
| Let's get started! Please structure your summary in JSON format. | |