--- library_name: transformers license: apache-2.0 base_model: - openbmb/MiniCPM5-1B language: - en - ru - zh pipeline_tag: text-generation tags: - director - supervisor - orchestrator - dubbing - dense - llama - llama-architecture - minicpm - minicpm5 - sft - lora-merged - bf16 - long-context - noesis - noesis-director - noesis-dhcf-fno - amaimedia - nt325 - vendored-internal - r-apache-clean - r-supervisor-always-connected - r-sft-director-no-think - r-llama-arch-standard - r-train-ckpt-50-resumable - r-vendored-internal - r-never-delete-without-explicit-consent inference: false datasets: - openbmb/Ultra-FineWeb - openbmb/Ultra-FineWeb-L3 - openbmb/UltraData-Math - openbmb/UltraData-SFT-2605 --- ## NOESIS / AMAImedia Released as part of the NOESIS Professional Multilingual Dubbing Automation Platform (framework: DHCF-FNO — Deterministic Hybrid Control Framework for Frozen Neural Operators). - **Founder:** Ilia Bolotnikov - **Organization:** AMAImedia.com - **X (Twitter):** [@AMAImediacom](https://x.com/AMAImediacom) - **LinkedIn:** [Ilia Bolotnikov](https://www.linkedin.com/in/ilia-bolotnikov) - **Telegram:** [@djbionicl](https://t.me/djbionicl) - **NOESIS version:** v16.1 - **Release date:** 2026-08-26 ================================================================================ NOESIS-Llama-1B-MiniCPM5-Orchestrator-Director-Supervisor-BF16 -- NOESIS Bundle README ================================================================================ > **NT-325 SFT director merge** — `peft.merge_and_unload` of the > NOESIS director-SFT LoRA (`LORA/nt325_sup_llama1b_minicpm5/adapter`) > over the NOESIS Llama-1B-MiniCPM5 Supervisor BF16 base. Role: > **director / supervisor / orchestrator** of the NOESIS dubbing > pipeline — dense Llama-arch small director, 128 K context, terse > decision-only output, no ``. NOESIS provenance ----------------- Bundle : NOESIS-Llama-1B-MiniCPM5-Orchestrator-Director-Supervisor-BF16 Parent bundle : NOESIS-3.5B-A0.5B-DUBBING-FILM Upstream chain : openbmb/MiniCPM5-1B (Apache-2.0) → NOESIS-Llama-1B-MiniCPM5-Supervisor-BF16 → NT-325 director-SFT LoRA merge → **this bundle** License : Apache License 2.0 (end-to-end) NOESIS variant : **BF16 merged director** (single dense Llama shard, ~2.16 GB). NT-325 director-SFT LoRA merged into the Llama-1B-MiniCPM5 Supervisor base via `peft.merge_and_unload`. NOESIS role : 1 B-tier director / supervisor / orchestrator with 128 K context (full cinema reel fits in one context). Sibling to the 0.8 B Qwopus director; both consume the same NOESIS-SUPERVISOR-DIRECTOR-FINAL-v3-97k.jsonl curriculum. Last updated : 2026-08-29 ================================================================================ NOESIS director ladder (small → large) ================================================================================ | Tier | Bundle | Role | |------|--------|------| | 0.8 B dense | ../NOESIS-Qwopus3.5-0.8B-v3-Orchestrator-Director-Supervisor-BF16 | Fast small director (sibling) | | **1.0 B dense** | **NOESIS-Llama-1B-MiniCPM5-Orchestrator-Director-Supervisor-BF16** ← *this bundle* | **Dense Llama-arch small director, 128 K context** | | 7.5 B MoE A1B | ../NOESIS-LFM2.5-7.5B-A1B-Orchestrator-Director-Supervisor-BF16 | Hybrid conv+attn MoE director | | 7.5 B MoE A2.5B | ../NOESIS-Mellum2-7.5B-A2.5B-Orchestrator-Director-Supervisor-BF16 | QWEN3MOE director | ================================================================================ Architecture / Files ================================================================================ | Property | Value | |----------|-------| | Architecture | `LlamaForCausalLM` (standard, R-LLAMA-ARCH-STANDARD) | | Total params | ~1.08 B | | Layers | 24 | | Attention heads | GQA 16 Q / 2 KV | | Context length | 131 072 (128 K) — full cinema reel fits in one context | | Precision | **BF16** (single shard) | | Weights file | `model.safetensors` (~2.16 GB) | | Tokenizer | MiniCPM5 tokenizer (`tokenizer.json` + `tokenizer_config.json`) | | Chat template | `chat_template.jinja` (director / no-think) | | Deploy quant | GGUF Q8_0 (sister deploy file, A/B clean) | ``` . ├── README.md ├── LICENSE # Apache 2.0 + NOESIS notice ├── model.safetensors # BF16 dense Llama (~2.16 GB) ├── config.json ├── generation_config.json ├── tokenizer.json ├── tokenizer_config.json ├── chat_template.jinja └── NOESIS_MERGE_MANIFEST.json # NT-325 merge provenance ``` ================================================================================ 3-way A/B — OLD upstream vs NEW NT-325 SFT vs GGUF Q8_0 (5 supervisor prompts) ================================================================================ Operator-verified 2026-06-06: | Variant | Garbage | Speed | Quality | |---|---|---|---| | **OLD upstream** (no NT-325 SFT) | 5 / 5 (all ``) | 15.1 tok/s | ❌ everything in reasoning trace | | **NEW NT-325 SFT** (this bundle) | **0 / 5 ✅** | **~17 tok/s** | ⚠️ no ``, but some answers weak (compress/expand direction sometimes wrong) | | **GGUF Q8_0** (sister deploy file) | 0 / 5 ✅ | — | ⚠️ prompt-echo, but `Paris`, works | → NT-325 SFT removed the `` leak ✅. Direction accuracy on isochrony decisions is **below the 0.8 B Qwopus sibling** — for hard routing / isochrony cases prefer the Qwopus 0.8 B director (sharper decisions); use this 1 B Llama bundle when **long context (>32 K, up to 128 K)** is required (full reel scope, multi-speaker QC). ================================================================================ How it was built — NT-325 director SFT merge ================================================================================ 1. **Base** — `NOESIS-Llama-1B-MiniCPM5-Supervisor-BF16` (Supervisor BF16 source, fine-tuned from `openbmb/MiniCPM5-1B`). 2. **Director-SFT LoRA** — `LORA/nt325_sup_llama1b_minicpm5/adapter`. Trained on `LORA/NOESIS-SUPERVISOR-DIRECTOR-FINAL-v3-97k.jsonl` (same curriculum as the 0.8 B / 7.5 B directors): routing / isochrony / gate decisions, canon-RZ22 stage transitions, no-`` enforcement. 3. **Merge** — `peft.merge_and_unload` → BF16 single shard (`scripts/nt325_merge_director_bf16.py`, 2026-06-06 21:29:21). 4. **GGUF Q8_0 deploy** — sister deploy file. A/B 0 / 5 garbage, prompt-echo visible but `Paris` probe passes. ================================================================================ VRAM and runtime ================================================================================ | Setup | VRAM | |-------|------| | BF16 native, no offload | ~2.6 GB | | BF16 + KV @ 4 K ctx | ~3.2 GB | | BF16 + KV @ 32 K ctx (typical cinema reel) | ~4.5 GB | | RTX 3060 6 GB | ✅ comfortable with sequential-swap policy | ================================================================================ NOESIS Sealed Rules ================================================================================ R-APACHE-CLEAN End-to-end Apache 2.0 lineage (MiniCPM5-1B → NOESIS Supervisor BF16 → NT-325 director SFT merge). Full upstream license text + NOTICE in `LICENSE`. R-SUPERVISOR-ALWAYS-CONNECTED Director / supervisor wired into the canon-RZ22 dub pipeline via `demo_server/qwopus_supervisor.py` (NEVER delete / orphan). R-SFT-DIRECTOR-NO-THINK (pattern) Production verdicts are flat, terse, decision-only. No `` / CoT leakage. `enable_thinking=False` in production. R-LLAMA-ARCH-STANDARD Standard `LlamaForCausalLM` — no custom kernels, no model-code fork. HF-loadable out of the box. R-TRAIN-CKPT-50-RESUMABLE NT-325 director SFT training: checkpoints every 50 steps, spot-instance resumable. R-VENDORED-INTERNAL Internal vendor copy inside parent NOESIS-3.5B-A0.5B-DUBBING-FILM. R-DUBBING-FILM-SCOPE (sealed 2026-04-29) NOESIS = professional audio dubbing. 1 B-tier director / supervisor of the canon-RZ22 dub pipeline. 128 K context for long-form multi-speaker reels. R-NOESIS-FINAL-ARTIFACT-PATHS (sealed 2026-05-27) Canonical path: models/llm/NOESIS-3.5B-A0.5B-DUBBING-FILM/ NOESIS-Llama-1B-MiniCPM5-Orchestrator-Director-Supervisor-BF16/ R-NEVER-DELETE-WITHOUT-EXPLICIT-CONSENT (sealed 2026-05-21) MUST NOT be deleted without explicit operator instruction "удали " / "delete ". 128 K-context director — long-reel QC role can't be filled by the 0.8 B Qwopus sibling. ================================================================================ Upstream Citation ================================================================================ Upstream model : https://huggingface.co/openbmb/MiniCPM5-1B License : Apache License 2.0 NT-325 SFT adapter source : LORA/nt325_sup_llama1b_minicpm5/adapter Training corpus : LORA/NOESIS-SUPERVISOR-DIRECTOR-FINAL-v3-97k.jsonl ```bibtex @article{minicpm4, title = {MiniCPM4: Ultra-efficient LLMs on end devices}, author = {MiniCPM Team}, journal= {arXiv preprint arXiv:2506.07900}, year = {2025} } ``` ================================================================================ NOESIS — Deterministic Hybrid Control Framework for Frozen Neural Operators (DHCF-FNO). Copyright (c) 2026 AMAImedia.com. All rights reserved. MiniCPM5-1B base weights © 2025 OpenBMB / MiniCPM Team, released under the Apache License 2.0. NOESIS Supervisor BF16 fine-tune + NT-325 director SFT merge + provenance © AMAImedia 2026 (NOESIS DHCF-FNO project, Apache-2.0).