---
license: apache-2.0
language:
- en
- es
- fr
- de
- it
- pt
- ru
- ar
- hi
- ko
- zh
library_name: transformers
base_model:
- arcee-ai/Trinity-Mini
base_model_relation: quantized
tags:
- moe
- nvfp4
- modelopt
- blackwell
- vllm
---
# Trinity Mini NVFP4
**This repository contains the NVFP4 quantized weights of Trinity-Mini for deployment on NVIDIA Blackwell GPUs.**
Trinity Mini is an Arcee AI 26B MoE model with 3B active parameters. It is the medium-sized model in our new Trinity family, a series of open-weight models for enterprise and tinkerers alike.
This model is tuned for reasoning, but in testing, it uses a similar total token count to competitive instruction-tuned models.
***
Trinity Mini is trained on 10T tokens gathered and curated through a key partnership with [Datology](https://www.datologyai.com/), building upon the excellent dataset we used on [AFM-4.5B](https://huggingface.co/arcee-ai/AFM-4.5B) with additional math and code.
Training was performed on a cluster of 512 H200 GPUs powered by [Prime Intellect](https://www.primeintellect.ai/) using HSDP parallelism.
More details, including key architecture decisions, can be found on our blog [here](https://www.arcee.ai/blog/the-trinity-manifesto)
***
## Model Details
* **Model Architecture:** AfmoeForCausalLM
* **Parameters:** 26B, 3B active
* **Experts:** 128 total, 8 active, 1 shared
* **Context length:** 128k
* **Training Tokens:** 10T
* **License:** [Apache 2.0](https://huggingface.co/arcee-ai/Trinity-Mini#license)
* **Recommended settings:**
* temperature: 0.15
* top_k: 50
* top_p: 0.75
* min_p: 0.06
***
## Benchmarks
## Quantization Details
- **Scheme:** NVFP4 (`nvfp4_mlp_only` — MLP/expert weights only, attention remains BF16)
- **Tool:** [NVIDIA ModelOpt](https://github.com/NVIDIA/Model-Optimizer)
- **Calibration:** 512 samples, seq_length=2048, all-expert calibration enabled
- **KV cache:** Not quantized
## Running with vLLM
Requires [vLLM](https://github.com/vllm-project/vllm) >= 0.18.0. Native FP4 compute requires Blackwell GPUs; older GPUs fall back to Marlin weight decompression automatically.
### Blackwell GPUs (B200/B300/GB300) — Docker (recommended)
```bash
docker run --runtime nvidia --gpus all -p 8000:8000 \
-v ~/.cache/huggingface:/root/.cache/huggingface \
vllm/vllm-openai:v0.18.0-cu130 \
arcee-ai/Trinity-Mini-NVFP4 \
--trust-remote-code \
--gpu-memory-utilization 0.90 \
--max-model-len 8192
```
### Hopper GPUs (H100/H200) and others
```bash
vllm serve arcee-ai/Trinity-Mini-NVFP4 \
--trust-remote-code \
--gpu-memory-utilization 0.90 \
--max-model-len 8192 \
--host 0.0.0.0 \
--port 8000
```
**Note (Blackwell pip installs):** If installing vLLM via pip on Blackwell rather than using Docker, native FP4 kernels may produce incorrect output due to package version mismatches. As a workaround, force the Marlin backend:
```bash
export VLLM_NVFP4_GEMM_BACKEND=marlin
vllm serve arcee-ai/Trinity-Mini-NVFP4 \
--trust-remote-code \
--moe-backend marlin \
--gpu-memory-utilization 0.90 \
--max-model-len 8192 \
--host 0.0.0.0 \
--port 8000
```
Marlin decompresses FP4 weights to BF16 for compute, providing the full memory compression benefit (~3.7× vs BF16) but not native FP4 compute speedup. On Hopper GPUs (H100/H200), Marlin is selected automatically and no extra flags are needed.
## License
Trinity-Mini-NVFP4 is released under the Apache-2.0 license.
