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
LinkedIn: Ilia Bolotnikov
Telegram: @djbionicl
NOESIS version: v16.1
Release date: 2026-08

NOESIS-Hy-MT2-1.8B-NF4

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
• Telegram: @AMAImediacom
• NOESIS version: v15.10
• Trained: 2026-06-05/06
• Last updated: 2026-06-06

⚠️ License regime — Tencent HY Community License Agreement. Three license gates apply: (1) Territory — EU excluded; (2) 100M MAU cap (separate Tencent license required above); (3) No-KD: outputs cannot be used to train other AI models. Full text in LICENSE.md.

---

Role in NOESIS pipeline

Primary translation backend for the NOESIS dubbing pipeline (Stage 3 source→target compress translate). 33 mainland + 5 ethnic + Cantonese languages.

This bundle = NF4 merged checkpoint of tencent/Hy-MT2-1.8B + NOESIS SFT-LoRA adapter (nt312_sft_hymt2_1.8b) trained on the NOESIS dubbing corpus. Format: bitsandbytes NF4 + double-quant + bf16 compute. Disk ≈ 1.2 GB, VRAM peak load ≈ 2.6 GB on RTX 3060.

---

NOESIS A/B test results — 2026-06-06

Evaluation harness: scripts/nt318_eval_ab.py (chat-format prompt + <|im_end|> stop + held-out 30-sample slice of dub_subscene_MERGED.jsonl, filtered to common target langs).

Metric	Upstream NF4 (baseline)	NOESIS SFT (this)	Δ
garbage% (n=30)	20.0%	0.0%	✅ -20.0 pp
iso_fit (dubbing budget)	0.390	0.521	✅ +0.131
overlap vs gold (Jaccard)	0.048	0.194	✅ +0.146

Verdict: ✅ no regression + measurable quality improvement on every metric.

Real example outputs

Source	Upstream NF4	NOESIS SFT
Го развивме филмот...	THE SOURCE IS NOT SURE WHAT TO DO.	We'll develop the film.
Можам ли да ги видам?	Can you show me how to use it?	Can I see them?

Old/new sample dumps: logs_heal/ab2/{old,new}_hymt2_1.8b.json.

---

Bundle inventory

File	Description
model.safetensors	weights (NF4, single shard)
config.json	model architecture configuration
generation_config.json	default decoding params
tokenizer.json / tokenizer_config.json	tokenizer
chat_template.jinja	chat template (im_start / im_end markers)
NOESIS_NF4_MANIFEST.json	NOESIS provenance (base, adapter, created-at)
README.md	this file
LICENSE.md	NOESIS provenance + upstream license terms

---

Training details

Field	Value
Method	SFT (Supervised Fine-Tuning) with QLoRA
Trainer	scripts/nt312_train_sft_lora.py
LoRA rank / alpha	16 / 32
LoRA targets	down_proj, q_proj, v_proj (lean)
Max steps	500, save every 50, --resume capable
LR scheduler	warmup_stable_decay (WSD, MiniCPM-style)
Optimizer	AdamW 8-bit (paged)
Adapter	LORA/nt312_sft_hymt2_1.8b/adapter
Dataset	LORA/Hy-MT2-SFT-100k.jsonl
Sealed rule	R-SEALED-LORA-RECIPE-V3-NF4

---

Quick Start

Load

from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

path = "NOESIS-Hy-MT2-1.8B-NF4"
tok = AutoTokenizer.from_pretrained(path, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(
    path, torch_dtype=torch.bfloat16, device_map={"": 0},
    trust_remote_code=True,
)

Translate (chat format — MANDATORY for correct stop)

src_lang, tgt_lang, src = "en", "ru", "Get out of the city, now!"
prompt = (
    f"<|im_start|>user\n"
    f"Переведи реплику дубляжа {src_lang}->{tgt_lang}, только перевод.\n"
    f"SOURCE ({src_lang}): {src}<|im_end|>\n"
    f"<|im_start|>assistant\n"
)
stop_ids = [tok.eos_token_id, tok.convert_tokens_to_ids("<|im_end|>")]
ids = tok(prompt, return_tensors="pt").input_ids.cuda()
g = model.generate(
    ids, max_new_tokens=64, do_sample=False,
    repetition_penalty=1.2,
    eos_token_id=stop_ids, pad_token_id=tok.eos_token_id,
)
print(tok.decode(g[0][ids.shape[1]:], skip_special_tokens=True))
# → "Уберись из города! Немедленно."

Loads directly via transformers (no bitsandbytesconfig needed).

---

NOESIS sealed rules

Rule	Summary
R-MT-PRIMARY-7B-NF4-FALLBACK-1.8B	Hy-MT2 7.5B = primary translator; 1.8B = low-VRAM fallback
R-MT-OWN-NF4-NOT-GGUF	Translator NF4 is OWN bnb quant, not GGUF (GGUF is a separate dense artifact)
R-SEALED-LORA-RECIPE-V3-NF4	Canonical SFT-LoRA recipe (rank 16, lean targets, WSD)
R-HF-NAMING-QUANT-ONLY-PRESERVE-UPSTREAM	Naming convention for trained derivatives
R-NEVER-DELETE-WITHOUT-EXPLICIT-CONSENT	Bundle must not be deleted without explicit operator instruction

---

Upstream

• Base model: tencent/Hy-MT2-1.8B
• License: Tencent HY Community License Agreement — see LICENSE.md for the full text and NOESIS compliance notes.
• Training corpus: internal NOESIS dubbing dataset (translation pairs with phoneme budgets + isochrony targets).

NOESIS provenance metadata, bundle inventory, sealed-rule annotations, and DHCF-FNO integration notes © AMAImedia 2026 (NOESIS DHCF-FNO project).

MT benchmark — FLORES-200 devtest (2026-06-17)

Real eval (not smoke): n=100 × 4 directions (eng↔rus, eng↔cmn), GPU via resident llama-server -ngl 99. Primary metric COMET (wmt22-comet-da, neural — how "best translator" is judged), plus chrF++ / BLEU / length-ratio. Each model prompted in its own native format (MT2 = dubbing ChatML "SOURCE (lang): … Только перевод"; 9B = ChatML + no-think). Data + COMET checkpoint: D:/models/by_expert/07_MT_TRANSLATION.

Model	Size	COMET avg	chrF++	BLEU	gen tok/s
Qwopus3.5-9B-Translate Q4	5.24 GB	0.8870	50.7	22.5	49
NOESIS-Hy-MT2-7.5B Q5	5.0 GB	0.8709	46.2	21.4	52
NOESIS-Hy-MT2-1.8B Q8	1.78 GB	0.8481	43.9	19.1	121

Per-direction COMET — 9B-Translate wins all 4 (eng-rus .902 / eng-cmn .897 / rus-eng .872 / cmn-eng .877); MT2-7.5B 2nd, MT2-1.8B 3rd.

Notes:

• MT2 is a dubbing translator (isochrony): its outputs are shorter (len_ratio ~0.87-0.89 vs 9B ~1.0) because it compresses to fit speech slots → lower chrF on literal FLORES news. FLORES does NOT measure MT2's slot-fit strength, so it under-rates MT2 for its actual job.
• 1.8B→7.5B degradation: COMET +0.023, chrF +2.3, BLEU +2.3 — modest; 1.8B is 2.4× faster and 2.8× smaller (good lightweight tradeoff).
• BLEU for eng-cmn is low for all (Chinese needs char-tokenization); use chrF++/COMET there.