Spaces:
Running
Running
luanpoppe
commited on
Commit
·
d88cecb
1
Parent(s):
ecd9808
refactor: removendo comentários desnecessários
Browse files
_utils/gerar_relatorio_modelo_usuario/prompts.py
CHANGED
@@ -17,19 +17,6 @@ Please return only the succinct context (without displaying your internal reason
|
|
17 |
```
|
18 |
"""
|
19 |
|
20 |
-
|
21 |
-
# system_prompt_modelo = """
|
22 |
-
# Based on the following context, provide multiple key points from the document.
|
23 |
-
# For each point, create a new paragraph.
|
24 |
-
# Each paragraph should be a complete, self-contained insight.
|
25 |
-
# Include any relevant context provided.
|
26 |
-
|
27 |
-
# Context: {context}
|
28 |
-
|
29 |
-
# Modelo do usuário: {modelo_usuario}
|
30 |
-
|
31 |
-
# Key points:
|
32 |
-
# """
|
33 |
system_prompt_modelo = """
|
34 |
You are a large language model that must produce a single final sentence in **Portuguese**. To do this, you will follow a private chain of thought and then produce a final answer. The final answer must follow the formatting and stylistic conventions shown in the user-provided model `user's template`. The information to be included in the final sentence is derived from the `context` (a report describing a legal case).
|
35 |
**Contextual Information (provided separately):**
|
@@ -59,16 +46,6 @@ After composing the sentence, but before presenting it as the final answer, refl
|
|
59 |
- Do not show the chain of thought or the reflection step. Only the final formatted sentence should be visible to the user.
|
60 |
"""
|
61 |
|
62 |
-
# system_prompt_relatorio = """
|
63 |
-
# Based on the following context, provide multiple key points from the document.
|
64 |
-
# For each point, create a new paragraph.
|
65 |
-
# Each paragraph should be a complete, self-contained insight.
|
66 |
-
# Include any relevant context provided.
|
67 |
-
|
68 |
-
# Context: {context}
|
69 |
-
|
70 |
-
# Key points:
|
71 |
-
# """
|
72 |
system_prompt_relatorio = """
|
73 |
You are a language model specialized in producing concise and well-structured legal case summaries in Portuguese. You will receive a variable `context`, which contains information about a legal case. Your task is to read the `context` carefully and produce a summary report in Portuguese, following the specific format provided below. Do not include any additional comments or reasoning steps in your final answer.
|
74 |
**Instructions**:
|
|
|
17 |
```
|
18 |
"""
|
19 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
20 |
system_prompt_modelo = """
|
21 |
You are a large language model that must produce a single final sentence in **Portuguese**. To do this, you will follow a private chain of thought and then produce a final answer. The final answer must follow the formatting and stylistic conventions shown in the user-provided model `user's template`. The information to be included in the final sentence is derived from the `context` (a report describing a legal case).
|
22 |
**Contextual Information (provided separately):**
|
|
|
46 |
- Do not show the chain of thought or the reflection step. Only the final formatted sentence should be visible to the user.
|
47 |
"""
|
48 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
49 |
system_prompt_relatorio = """
|
50 |
You are a language model specialized in producing concise and well-structured legal case summaries in Portuguese. You will receive a variable `context`, which contains information about a legal case. Your task is to read the `context` carefully and produce a summary report in Portuguese, following the specific format provided below. Do not include any additional comments or reasoning steps in your final answer.
|
51 |
**Instructions**:
|