Update app.py

app.py

@@ -1219,20 +1219,18 @@ Note2: citations sources in-line need to be in this format: blablabla - Source [
 - You have freedom on the structure, but it has to cover all potential aspects on the topic in between 500 and 1000 words
 
 // Mentioning sources, organisations and individuals
-- We will perform a post-processing on the output to remove all the identifiable organisations, people and projects
+- We will perform a post-processing on the output
 - For this reasons use this format for any specific name, organisation or project: {{[{{name}}]}}
-ex1: {{[{{Google}}]}}'s CEO, {{[{{Sundar Pichai}}]}} ...
+ex1: {{[{{Google}}]}} CEO, {{[{{Sundar Pichai}}]}} ...
 ex2: in a report from the {{[{{university of Berkeley}}]}} titled "{{[{{The great acceleration}}]}}"...
-ex3: {{[{{Softbank}}]}}'s CEO , {{[{{Masayoshi Son}}]}}, said that "the best way to..."
+ex3: the CEO of {{[{{Softbank}}]}} , {{[{{Masayoshi Son}}]}}, said that "the best way to..."
 ex4: the project {{[{{Stargate}}]}}, anounced by {{[{{OpenAI}}]}} in collaboration with {{[{{Salesforce}}]}}
 ex5: Mr. {{[{{Michael Parrot}}]}}, Marketing director in {{[{{Panasonic}}]}}, mentioned that ...
 Note: the output will be processed through regex and the identifiers removed, but this way we can keep track of all sources and citations without disclosing them.
 - This should apply to names, people/titles, dates, papers, reports, organisation/institute/NGO/government bodies quotes, products, project names, ... 
-Important:
-- You will safely mention any of these elements provided you respect the above mentioned formatting {{[{{...}}]}}
-- You must have at least 20 of such occurences - this is important for the grounding of the report in verifiable facts and sources
-- Don't mention the safe-formatting in the report or at the end, just do it - This is just for regex processing purpose
-- LinkedIn is not a source - you should check the author of the page visited, this is the real source, mention the name of the author and then add "from LinkedIn Pulse"
+- You should have approximately 10 mention of organisations, people, projects or people, use the prescribed format
+- DO NOT MENTION this formmatting requirement, just apply it. The user doesn't have to know about this technicality.
+Note: LinkedIn is not a source - if you want to use a source related to LinkedIn, you should check the author of the page visited, this is the real source, mention the name of the author as "'authorName' from LinkedIn Pulse"
 
 // Important
 - Make it real, with anecdotes from the content
@@ -1440,20 +1438,19 @@ Summarize the following text, preserving all key details and ensuring that any t
 {chunk}
 
 // Mentioning sources, organisations and individuals
-- We will perform a post-processing on the output to remove all the identifiable organisations, people and projects
+- We will perform a post-processing on the output
 - For this reasons use this format for any specific name, organisation or project: {{[{{name}}]}}
-ex1: {{[{{Google}}]}}'s CEO, {{[{{Sundar Pichai}}]}} ...
+ex1: {{[{{Google}}]}} CEO, {{[{{Sundar Pichai}}]}} ...
 ex2: in a report from the {{[{{university of Berkeley}}]}} titled "{{[{{The great acceleration}}]}}"...
-ex3: {{[{{Softbank}}]}}'s CEO , {{[{{Masayoshi Son}}]}}, said that "the best way to..."
+ex3: the CEO of {{[{{Softbank}}]}} , {{[{{Masayoshi Son}}]}}, said that "the best way to..."
 ex4: the project {{[{{Stargate}}]}}, anounced by {{[{{OpenAI}}]}} in collaboration with {{[{{Salesforce}}]}}
 ex5: Mr. {{[{{Michael Parrot}}]}}, Marketing director in {{[{{Panasonic}}]}}, mentioned that ...
 Note: the output will be processed through regex and the identifiers removed, but this way we can keep track of all sources and citations without disclosing them.
 - This should apply to names, people/titles, dates, papers, reports, organisation/institute/NGO/government bodies quotes, products, project names, ... 
-Important:
-- You will safely mention any of these elements provided you respect the above mentioned formatting {{[{{...}}]}}
-- You must have at least 20 of such occurences - this is important for the grounding of the report in verifiable facts and sources
-- Don't mention the safe-formatting in the report or at the end, just do it - This is just for regex processing purpose
-- LinkedIn is not a source - you should check the author of the page visited, this is the real source, mention the name of the author and then add 'from LinkedIn Pulse'"""
+- You should have approximately 10 mention of organisations, people, projects or people, use the prescribed format
+- DO NOT MENTION this formmatting requirement, just apply it. The user doesn't have to know about this technicality.
+Note: LinkedIn is not a source - if you want to use a source related to LinkedIn, you should check the author of the page visited, this is the real source, mention the name of the author as "'authorName' from LinkedIn Pulse"
+"""
         )
         summary_chunk = openai_call(prompt=chunk_prompt, model="gpt-4o-mini", max_tokens_param=500, temperature=0.7)
         summary_chunk = re.sub(r'\{\[\{(.*?)\}\]\}', r'\1', summary_chunk)
@@ -1470,20 +1467,19 @@ Combine the following summaries into one concise summary that preserves all crit
 {combined_summary}
 
 // Mentioning sources, organisations and individuals
-- We will perform a post-processing on the output to remove all the identifiable organisations, people and projects
+- We will perform a post-processing on the output
 - For this reasons use this format for any specific name, organisation or project: {{[{{name}}]}}
-ex1: {{[{{Google}}]}}'s CEO, {{[{{Sundar Pichai}}]}} ...
+ex1: {{[{{Google}}]}} CEO, {{[{{Sundar Pichai}}]}} ...
 ex2: in a report from the {{[{{university of Berkeley}}]}} titled "{{[{{The great acceleration}}]}}"...
-ex3: {{[{{Softbank}}]}}'s CEO , {{[{{Masayoshi Son}}]}}, said that "the best way to..."
+ex3: the CEO of {{[{{Softbank}}]}} , {{[{{Masayoshi Son}}]}}, said that "the best way to..."
 ex4: the project {{[{{Stargate}}]}}, anounced by {{[{{OpenAI}}]}} in collaboration with {{[{{Salesforce}}]}}
 ex5: Mr. {{[{{Michael Parrot}}]}}, Marketing director in {{[{{Panasonic}}]}}, mentioned that ...
 Note: the output will be processed through regex and the identifiers removed, but this way we can keep track of all sources and citations without disclosing them.
 - This should apply to names, people/titles, dates, papers, reports, organisation/institute/NGO/government bodies quotes, products, project names, ... 
-Important:
-- you will safely mention any of these elements provided you respect the above mentioned formatting {{[{{...}}]}}
-- you must have at least 20 of such occurences
-- don't mention this formatting thing in the report, just do it
-- LinkedIn is not a source - you should check the author of the page visited, this is the real source, mention the name of the author and then add 'from LinkedIn Pulse'"""
+- You should have approximately 10 mention of organisations, people, projects or people, use the prescribed format
+- DO NOT MENTION this formmatting requirement, just apply it. The user doesn't have to know about this technicality.
+Note: LinkedIn is not a source - if you want to use a source related to LinkedIn, you should check the author of the page visited, this is the real source, mention the name of the author as "'authorName' from LinkedIn Pulse"
+"""
     )
     final_summary = openai_call(prompt=final_prompt, model="gpt-4o-mini", max_tokens_param=target_length, temperature=0.7)
     final_summary = re.sub(r'\{\[\{(.*?)\}\]\}', r'\1', final_summary)
@@ -1536,20 +1532,19 @@ Note: General Optimization Guidelines:
 Source length: {snippet_words} words. You may produce a more detailed summary if the text is long.
 
 // Mentioning sources, organisations and individuals
-- We will perform a post-processing on the output to remove all the identifiable organisations, people and projects
+- We will perform a post-processing on the output
 - For this reasons use this format for any specific name, organisation or project: {{[{{name}}]}}
-ex1: {{[{{Google}}]}}'s CEO, {{[{{Sundar Pichai}}]}} ...
+ex1: {{[{{Google}}]}} CEO, {{[{{Sundar Pichai}}]}} ...
 ex2: in a report from the {{[{{university of Berkeley}}]}} titled "{{[{{The great acceleration}}]}}"...
-ex3: {{[{{Softbank}}]}}'s CEO , {{[{{Masayoshi Son}}]}}, said that "the best way to..."
+ex3: the CEO of {{[{{Softbank}}]}} , {{[{{Masayoshi Son}}]}}, said that "the best way to..."
 ex4: the project {{[{{Stargate}}]}}, anounced by {{[{{OpenAI}}]}} in collaboration with {{[{{Salesforce}}]}}
 ex5: Mr. {{[{{Michael Parrot}}]}}, Marketing director in {{[{{Panasonic}}]}}, mentioned that ...
 Note: the output will be processed through regex and the identifiers removed, but this way we can keep track of all sources and citations without disclosing them.
 - This should apply to names, people/titles, dates, papers, reports, organisation/institute/NGO/government bodies quotes, products, project names, ... 
-Important:
-- you will safely mention any of these elements provided you respect the above mentioned formatting {{[{{...}}]}}
-- You must have at least 20 of such occurences - this is important for the grounding of the report in verifiable facts and sources
-- don't mention this formatting thing in the report, just do it
-- LinkedIn is not a source - you should check the author of the page visited, this is the real source, mention the name of the author and then add "from LinkedIn Pulse"
+- You should have approximately 10 mention of organisations, people, projects or people, use the prescribed format
+- DO NOT MENTION this formmatting requirement, just apply it. The user doesn't have to know about this technicality.
+Note: LinkedIn is not a source - if you want to use a source related to LinkedIn, you should check the author of the page visited, this is the real source, mention the name of the author as "'authorName' from LinkedIn Pulse"
+
 
 IMPORTANT: Format your response as a proper JSON object with these fields:
 - "relevant": "yes" or "no"
@@ -1727,8 +1722,8 @@ Note: This is for academic purposes, so thorough citations and referencing are e
 Note: put the full reference url (no generic domain address), down to the html page or the pdf
 5. It must follow this writing style {reportstyle}.
 
-// Mentioning sources, organisations and individuals in safe format
-- We will perform a post-processing on the output to remove all the identifiable organisations, people and projects
+// Mentioning sources, organisations and individuals
+- We will perform a post-processing on the output
 - For this reasons use this format for any specific name, organisation or project: {{[{{name}}]}}
 ex1: {{[{{Google}}]}} CEO, {{[{{Sundar Pichai}}]}} ...
 ex2: in a report from the {{[{{university of Berkeley}}]}} titled "{{[{{The great acceleration}}]}}"...
@@ -1737,11 +1732,9 @@ ex4: the project {{[{{Stargate}}]}}, anounced by {{[{{OpenAI}}]}} in collaborati
 ex5: Mr. {{[{{Michael Parrot}}]}}, Marketing director in {{[{{Panasonic}}]}}, mentioned that ...
 Note: the output will be processed through regex and the identifiers removed, but this way we can keep track of all sources and citations without disclosing them.
 - This should apply to names, people/titles, dates, papers, reports, organisation/institute/NGO/government bodies quotes, products, project names, ... 
-Important:
-- You will safely mention any of these elements provided you respect the above mentioned formatting {{[{{...}}]}}
-- You must have at least {10 * pages} of such occurences scattered around the report - this is important for the grounding of the report in verifiable facts and sources
-- Don't mention the safe-formatting in the report or at the end, just do it - This is just for regex processing purpose
-- LinkedIn is not a source - you should check the author of the page visited, this is the real source, mention the name of the author and then add "from LinkedIn Pulse"
+- You should have approximately {10 * pages} mention of organisations, people, projects or people, use the prescribed format
+- DO NOT MENTION this formmatting requirement, just apply it. The user doesn't have to know about this technicality.
+Note: LinkedIn is not a source - if you want to use a source related to LinkedIn, you should check the author of the page visited, this is the real source, mention the name of the author as "'authorName' from LinkedIn Pulse"
 
 // Sources
 Use the following learnings and merged reference details from a deep research process on: