mirror of
https://github.com/onestardao/WFGY.git
synced 2026-04-30 12:39:55 +00:00
87ff7f2fcf
- Add MiniMax-M2.7 and MiniMax-M2.7-highspeed to model list - Set MiniMax-M2.7 as default model - Keep all previous models as alternatives
163 lines
11 KiB
Markdown
163 lines
11 KiB
Markdown
# MiniMax — Guardrails and Fix Patterns
|
||
|
||
<details>
|
||
<summary><strong>🧭 Quick Return to Map</strong></summary>
|
||
|
||
<br>
|
||
|
||
> You are in a sub-page of **LLM_Providers**.
|
||
> To reorient, go back here:
|
||
>
|
||
> - [**LLM_Providers** — model vendors and deployment options](./README.md)
|
||
> - [**WFGY Global Fix Map** — main Emergency Room, 300+ structured fixes](../README.md)
|
||
> - [**WFGY Problem Map 1.0** — 16 reproducible failure modes](../../README.md)
|
||
>
|
||
> Think of this page as a desk within a ward.
|
||
> If you need the full triage and all prescriptions, return to the Emergency Room lobby.
|
||
</details>
|
||
|
||
|
||
Use this page when failures look provider specific on MiniMax models (MiniMax-M2.7, MiniMax-M2.7-highspeed, MiniMax-M2.5, MiniMax-M2.5-highspeed). MiniMax-M2.7 is the latest flagship model with enhanced reasoning and coding capabilities. Examples include temperature rejection at zero, tool-call JSON drift in long 204K-context windows, Chinese tokenizer similarity mismatches, or streaming stalls under high concurrency. Each fix maps back to WFGY pages so you can verify with measurable targets.
|
||
|
||
**Core acceptance**
|
||
|
||
- ΔS(question, retrieved) ≤ 0.45
|
||
- Coverage ≥ 0.70 for the target section
|
||
- λ remains convergent across 3 paraphrases
|
||
|
||
---
|
||
|
||
## Open these first
|
||
|
||
- Visual map and recovery: [RAG Architecture & Recovery](https://github.com/onestardao/WFGY/blob/main/ProblemMap/rag-architecture-and-recovery.md)
|
||
- End-to-end knobs: [Retrieval Playbook](https://github.com/onestardao/WFGY/blob/main/ProblemMap/retrieval-playbook.md)
|
||
- Why this snippet: [Retrieval Traceability](https://github.com/onestardao/WFGY/blob/main/ProblemMap/retrieval-traceability.md)
|
||
- Ordering control: [Rerankers](https://github.com/onestardao/WFGY/blob/main/ProblemMap/rerankers.md)
|
||
- Embedding vs meaning: [Embedding ≠ Semantic](https://github.com/onestardao/WFGY/blob/main/ProblemMap/embedding-vs-semantic.md)
|
||
- Hallucination and chunk boundaries: [Hallucination](https://github.com/onestardao/WFGY/blob/main/ProblemMap/hallucination.md)
|
||
- Long threads and memory: [Context Drift](https://github.com/onestardao/WFGY/blob/main/ProblemMap/context-drift.md), [Entropy Collapse](https://github.com/onestardao/WFGY/blob/main/ProblemMap/entropy-collapse.md), [Memory Coherence](https://github.com/onestardao/WFGY/blob/main/ProblemMap/memory-coherence.md)
|
||
- Logic collapse and recovery: [Logic Collapse](https://github.com/onestardao/WFGY/blob/main/ProblemMap/logic-collapse.md)
|
||
- Snippet and citation schema: [Data Contracts](https://github.com/onestardao/WFGY/blob/main/ProblemMap/data-contracts.md)
|
||
- Patterns: [Query Parsing Split](https://github.com/onestardao/WFGY/blob/main/ProblemMap/patterns/pattern_query_parsing_split.md), [Vectorstore Fragmentation](https://github.com/onestardao/WFGY/blob/main/ProblemMap/patterns/pattern_vectorstore_fragmentation.md), [Hallucination Re-entry](https://github.com/onestardao/WFGY/blob/main/ProblemMap/patterns/pattern_hallucination_reentry.md)
|
||
- Ops: [Live Monitoring](https://github.com/onestardao/WFGY/blob/main/ProblemMap/ops/live_monitoring_rag.md), [Debug Playbook](https://github.com/onestardao/WFGY/blob/main/ProblemMap/ops/debug_playbook.md)
|
||
- Multi-agent overview: [Multi-Agent Problems](https://github.com/onestardao/WFGY/blob/main/ProblemMap/Multi-Agent_Problems.md)
|
||
|
||
---
|
||
|
||
## Fix in 60 seconds
|
||
|
||
1) **Measure ΔS**
|
||
- Compute ΔS(question, retrieved) and ΔS(retrieved, expected anchor).
|
||
- Thresholds: stable < 0.40, transitional 0.40–0.60, risk ≥ 0.60.
|
||
|
||
2) **Probe with λ_observe**
|
||
- Vary k = {5, 10, 20}. Flat high curve suggests index or metric mismatch.
|
||
- Reorder prompt headers. If ΔS spikes, lock the schema.
|
||
|
||
3) **Apply the module**
|
||
- Retrieval drift → BBMC + Data Contracts.
|
||
- Reasoning collapse → BBCR bridge + BBAM variance clamp.
|
||
- Dead ends in long runs → BBPF alternate path.
|
||
|
||
4) **Provider knobs to check first**
|
||
- Temperature must be strictly greater than 0. Use 0.01 as the near-deterministic default.
|
||
- OpenAI-compatible endpoint (`https://api.minimax.io/v1`) — confirm base URL is set correctly.
|
||
- Structured output mode on and schema fixed.
|
||
- Tool use set to serial if parallel calls cross-talk.
|
||
|
||
5) **Verify**
|
||
- Three paraphrases hold the same citations.
|
||
- λ convergent across seeds.
|
||
- E_resonance flat on long replies.
|
||
|
||
---
|
||
|
||
## Typical breakpoints and the right fix
|
||
|
||
- **Temperature = 0 rejected, runs fail before any output**
|
||
MiniMax requires temperature in (0.0, 1.0]. Setting exactly 0 raises an API error. Use 0.01 for near-deterministic behavior. If your framework hardcodes `temperature=0` for evals, patch it before blaming retrieval. This is a config-level issue, not a semantic one.
|
||
|
||
- **Tool call JSON drifts or fields missing on long contexts**
|
||
The 204K context window allows very large prompts, but deep context can cause tool-call schema drift. Lock a strict IO header and cite the schema: [Data Contracts](https://github.com/onestardao/WFGY/blob/main/ProblemMap/data-contracts.md). Add trace tags in the prompt then verify: [Retrieval Traceability](https://github.com/onestardao/WFGY/blob/main/ProblemMap/retrieval-traceability.md). If agents are orchestrating, isolate boundaries: [Agent Boundary Design](https://github.com/onestardao/WFGY/blob/main/ProblemMap/agent-boundary-design.md), [Agent Consensus](https://github.com/onestardao/WFGY/blob/main/ProblemMap/agent-consensus-protocols.md).
|
||
|
||
- **Chinese tokenizer quirks change similarity despite high cosine**
|
||
Treat as metric mismatch. Use [Embedding ≠ Semantic](https://github.com/onestardao/WFGY/blob/main/ProblemMap/embedding-vs-semantic.md) and add BM25 fallback in the [Retrieval Playbook](https://github.com/onestardao/WFGY/blob/main/ProblemMap/retrieval-playbook.md). Then re-rank with [Rerankers](https://github.com/onestardao/WFGY/blob/main/ProblemMap/rerankers.md) and anchor citations via [Retrieval Traceability](https://github.com/onestardao/WFGY/blob/main/ProblemMap/retrieval-traceability.md).
|
||
|
||
- **Safety filter strips citations or tool arguments**
|
||
Move citation text to a dedicated field in the schema and reference with IDs. See [Retrieval Traceability](https://github.com/onestardao/WFGY/blob/main/ProblemMap/retrieval-traceability.md). If the model "bluffs" when filtered, apply controls in [Bluffing](https://github.com/onestardao/WFGY/blob/main/ProblemMap/bluffing.md).
|
||
|
||
- **Long chat melts down after filling the 204K window**
|
||
MiniMax models (M2.7, M2.5) support up to 204K tokens. Entropy collapse can still occur at the tail of long sessions. Cut context windows at stable joins and verify with [Context Drift](https://github.com/onestardao/WFGY/blob/main/ProblemMap/context-drift.md) and [Entropy Collapse](https://github.com/onestardao/WFGY/blob/main/ProblemMap/entropy-collapse.md). If replies flip across turns, check [Memory Desync](https://github.com/onestardao/WFGY/blob/main/ProblemMap/patterns/pattern_memory_desync.md).
|
||
|
||
- **OpenAI SDK client misconfiguration**
|
||
MiniMax uses an OpenAI-compatible API. When using the OpenAI SDK, set `base_url="https://api.minimax.io/v1"` and pass your MiniMax API key. Common pitfall: forgetting to change the base URL or passing the wrong key results in auth errors that look like model failures.
|
||
|
||
- **Hybrid retrieval (HyDE + BM25) underperforms**
|
||
Look for query splits in [Pattern: Query Parsing Split](https://github.com/onestardao/WFGY/blob/main/ProblemMap/patterns/pattern_query_parsing_split.md). Align the query parse and re-rank.
|
||
|
||
- **Non-English corpus drifts**
|
||
Follow the [Multilingual Guide](https://github.com/onestardao/WFGY/blob/main/ProblemMap/multilingual-guide.md). Normalize punctuation and numerals in chunking and traceability.
|
||
|
||
---
|
||
|
||
## Copy-paste prompt
|
||
|
||
```
|
||
|
||
I uploaded TXT OS and the WFGY Problem Map files.
|
||
|
||
My MiniMax bug:
|
||
• symptom: \[brief]
|
||
• traces: \[ΔS(question,retrieved)=..., ΔS(retrieved,anchor)=..., λ states]
|
||
|
||
Tell me:
|
||
|
||
1. which layer is failing and why,
|
||
2. which exact fix page to open from this repo,
|
||
3. the minimal steps to push ΔS ≤ 0.45 and keep λ convergent,
|
||
4. how to verify the fix with a reproducible test.
|
||
|
||
Use BBMC/BBPF/BBCR/BBAM where relevant.
|
||
|
||
```
|
||
|
||
---
|
||
|
||
## Escalate when
|
||
|
||
- First call after deploy fails or tools fire before data is ready. See [Pre-Deploy Collapse](https://github.com/onestardao/WFGY/blob/main/ProblemMap/predeploy-collapse.md) and [Bootstrap Ordering](https://github.com/onestardao/WFGY/blob/main/ProblemMap/bootstrap-ordering.md).
|
||
- Deadlocks or version skew in prod. See [Deployment Deadlock](https://github.com/onestardao/WFGY/blob/main/ProblemMap/deployment-deadlock.md).
|
||
|
||
---
|
||
|
||
### 🔗 Quick-Start Downloads (60 sec)
|
||
|
||
| Tool | Link | 3-Step Setup |
|
||
|------|------|--------------|
|
||
| **WFGY 1.0 PDF** | [Engine Paper](https://github.com/onestardao/WFGY/blob/main/I_am_not_lizardman/WFGY_All_Principles_Return_to_One_v1.0_PSBigBig_Public.pdf) | 1️⃣ Download · 2️⃣ Upload to your LLM · 3️⃣ Ask "Answer using WFGY + <your question>" |
|
||
| **TXT OS (plain-text OS)** | [TXTOS.txt](https://github.com/onestardao/WFGY/blob/main/OS/TXTOS.txt) | 1️⃣ Download · 2️⃣ Paste into any LLM chat · 3️⃣ Type "hello world" — OS boots instantly |
|
||
|
||
---
|
||
|
||
<!-- WFGY_FOOTER_START -->
|
||
|
||
### Explore More
|
||
|
||
| Layer | Page | What it's for |
|
||
| --- | --- | --- |
|
||
| ⭐ Proof | [WFGY Recognition Map](/recognition/README.md) | External citations, integrations, and ecosystem proof |
|
||
| ⚙️ Engine | [WFGY 1.0](/legacy/README.md) | Original PDF tension engine and early logic sketch (legacy reference) |
|
||
| ⚙️ Engine | [WFGY 2.0](/core/README.md) | Production tension kernel for RAG and agent systems |
|
||
| ⚙️ Engine | [WFGY 3.0](/TensionUniverse/EventHorizon/README.md) | TXT based Singularity tension engine (131 S class set) |
|
||
| 🗺️ Map | [Problem Map 1.0](/ProblemMap/README.md) | Flagship 16 problem RAG failure taxonomy and fix map |
|
||
| 🗺️ Map | [Problem Map 2.0](/ProblemMap/wfgy-rag-16-problem-map-global-debug-card.md) | Global Debug Card for RAG and agent pipeline diagnosis |
|
||
| 🗺️ Map | [Problem Map 3.0](/ProblemMap/wfgy-ai-problem-map-troubleshooting-atlas.md) | Global AI troubleshooting atlas and failure pattern map |
|
||
| 🧰 App | [TXT OS](/OS/README.md) | .txt semantic OS with fast bootstrap |
|
||
| 🧰 App | [Blah Blah Blah](/OS/BlahBlahBlah/README.md) | Abstract and paradox Q&A built on TXT OS |
|
||
| 🧰 App | [Blur Blur Blur](/OS/BlurBlurBlur/README.md) | Text to image generation with semantic control |
|
||
| 🏡 Onboarding | [Starter Village](/StarterVillage/README.md) | Guided entry point for new users |
|
||
|
||
If this repository helped, starring it improves discovery so more builders can find the docs and tools.
|
||
[](https://github.com/onestardao/WFGY)
|
||
|
||
<!-- WFGY_FOOTER_END -->
|