README.md

---
license: apache-2.0
language:
- en
- it
base_model: nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16
library_name: transformers
pipeline_tag: text-generation
tags:
- nemotron_h
- sync2
- fractal-rl
- cognitive-behaviors
- self-improving
- thesia
- alignment
- merged
- nemotron-h
- conversational
- custom_code
- endpoints-ready
- tgi-compatible
- vllm-compatible
inference:
  parameters:
    temperature: 0.7
    top_p: 0.9
    max_new_tokens: 512
    do_sample: true
widget:
- messages:
  - role: system
    content: You are sync2, a reasoner trained with Fractal RL.
  - role: user
    content: Spiega il principio di Ollivier-Ricci in 3 punti.
  example_title: Reasoning · IT
- messages:
  - role: user
    content: Write a Python function that computes Ollivier-Ricci curvature on a graph.
  example_title: Code · EN
model-index:
- name: Lorenzob/synch-2-merged
  results:
  - task:
      type: text-generation
      name: Cognitive Reasoning · Fractal-RL Composite
    dataset:
      type: internal
      name: Synch2 Internal Eval (private)
    metrics:
    - type: fractal_rl_reward
      value: 0.2
      name: Best Fractal-RL Composite Reward
extra_gated_prompt: Accept the Apache-2.0 license and the NVIDIA Nemotron-3 base license.
  The model embeds Lorenzo Bernardini's Fractal-RL / THESIA research; cite arxiv:2503.01307
  for cognitive-behaviors methodology.
extra_gated_fields:
  Affiliation: text
  Country: country
  Intended use: text
---
# Lorenzob/synch-2-merged

**Full merged model**: `nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16` (120B MoE · 12B active) **+**
`Lorenzob/synch-2` (v11 APOGEO LoRA · best reward `+0.200`).

Drop-in compatibile con: **HF Dedicated Endpoints**, **TGI**, **vLLM**,
**HF Inference API**, Together AI, Modal, RunPod, Replicate.

## Quickstart · Dedicated Endpoint
```python
from huggingface_hub import InferenceClient
client = InferenceClient("Lorenzob/synch-2-merged", token="<HF_TOKEN>")
out = client.chat_completion(
    messages=[
        {"role": "user",
         "content": "Compute the Ollivier-Ricci curvature of K_5."},
    ],
    max_tokens=512, temperature=0.7,
)
print(out.choices[0].message.content)
```

## Quickstart · TGI (Text Generation Inference)
```bash
docker run --gpus all --shm-size 1g -p 8080:80 \
    -v $PWD/data:/data \
    ghcr.io/huggingface/text-generation-inference:latest \
    --model-id Lorenzob/synch-2-merged --trust-remote-code \
    --num-shard 4 --max-input-length 4096 --max-total-tokens 8192
```

## Quickstart · vLLM
```bash
python -m vllm.entrypoints.openai.api_server \
    --model Lorenzob/synch-2-merged --trust-remote-code \
    --dtype bfloat16 --tensor-parallel-size 4 \
    --max-model-len 8192
```

## Quickstart · Locale (transformers full-weights)
```python
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

tok = AutoTokenizer.from_pretrained(
    "Lorenzob/synch-2-merged", trust_remote_code=True,
)
model = AutoModelForCausalLM.from_pretrained(
    "Lorenzob/synch-2-merged",
    torch_dtype=torch.bfloat16,
    device_map="auto",
    trust_remote_code=True,
)
```

## Suggested Hardware
AWS · 4x H100 80GB (consigliato) — accetta 8x A100 80GB — pesi BF16 totali ~245 GB,
50 shard `model-XXXXX-of-00050.safetensors`.

## Merge Details
- **Base**: NVIDIA Nemotron-3-Super-120B-A12B-BF16 (120B MoE, 12B active)
- **Adapter**: `Lorenzob/synch-2` (v11 APOGEO, best reward +0.200)
- **LoRA rank**: 64 · **LoRA alpha**: 32
- **Merge type**: weight addition (peft `merge_and_unload`)
- **Attention impl**: SDPA (default), FlashAttention-2 supported

## Governance
Vedi i documenti dedicati nel repo:
- [`bias.md`](bias.md) — bias analysis
- [`safety.md`](safety.md) — safety considerations
- [`privacy.md`](privacy.md) — privacy implications
- [`explainability.md`](explainability.md) — interpretability notes
- [`accuracy_chart.png`](accuracy_chart.png) — eval results visual

## Attribution
- Cognitive behaviors: Gandhi et al. 2025 ([arXiv:2503.01307](https://arxiv.org/abs/2503.01307))
- Self-improving reasoner: [karpathy/nanochat](https://github.com/karpathy/nanochat)
- Fractal RL · LCTR · THESIA: Lorenzo Bernardini publications.

## License
Apache-2.0 (this merged model) · NVIDIA Nemotron-3 license
applies to the underlying base weights distribution.