Reply to thread

Message: [QUOTE="Munyaradzi Mafaro, post: 68665, member: 636"] You are absolutely correct. My previous advice was for a [I]simple[/I] task, but rewriting a 2000-word article is a [B]complex task[/B]. Your finding that "Thinking Mode" is better is spot-on. For a long-form rewrite, the model isn't just rephrasing sentences. It has to perform a multi-step reasoning process: [LIST=1] [*]Read and understand the entire ~2000-word article. [*]Deconstruct its core arguments, structure, and tone. [*]Create a [I]plan[/I] for the rewritten version. [*]Execute that plan, rewriting section by section while maintaining coherence and consistency with the [I]other[/I] sections. [/LIST] This is precisely what "Thinking Mode" is built for. Using "Non-Thinking Mode" would likely result in a shallow or incoherent rewrite. [HEADING=2]The Token Math[/HEADING] [LIST] [*][B]Your Input:[/B] 2,000 words is roughly [B]2,700 - 3,000 tokens[/B]. [*][B]Your Output:[/B] The rewrite will also be around [B]3,000 tokens[/B]. [*][B]The "Thinking":[/B] The budget needs to be large enough for the model to "think" about the 3,000-token input and formulate its 3,000-token output plan. [/LIST] [HEADING=2]Recommended "Thinking Budget"[/HEADING] Based on official Qwen documentation for complex tasks, here is a tiered recommendation. [HEADING=3]1. Balanced Recommendation (Start Here)[/HEADING] [LIST] [*][B]Budget: 8,192 tokens[/B] [/LIST] This is a very solid and safe starting point. It gives the model ample "scratchpad" space to plan the rewrite of a 3,000-token article. In Qwen's own testing, a budget of 8,192 tokens was used for long-context "needle-in-a-haystack" tasks, which are similarly complex. [HEADING=3]2. High-Quality / Max-Effort[/HEADING] [LIST] [*][B]Budget: 16,384 tokens or 32,768 tokens[/B] [/LIST] If you find the 8k budget rewrite is still not detailed enough, or if the article is highly technical, increasing the budget will give the model more room for in-depth analysis and planning. The Qwen team officially recommends an [I]output length[/I] of 32,768 tokens for "most queries" and "highly complex problems," which implies that a very large thinking budget is supported and encouraged. [HEADING=3]3. Faster (Riskier)[/HEADING] [LIST] [*][B]Budget: 4,096 tokens[/B] [/LIST] You can try this if you need a faster result, but it's a "tight" budget. The risk is that the model's planning will be cut short, leading to a rewrite that "forgets" the plan halfway through or fails to maintain a consistent tone. [B]My recommendation:[/B] [B]Start with a thinking budget of 8,192 tokens.[/B] If the quality is perfect, you have your answer. If it seems rushed, increase the budget to 16,384. [/QUOTE]

Name