README.md

---
license: apache-2.0
language:
- en
library_name: transformers
tags:
- text-generation
- causal-lm
- transformer
- argonne
- instruct
- pretrained
pipeline_tag: text-generation
---

# Argonne 2.5-instruct

Argonne 2.5-instruct starts from the [Argonne 2.5-base](https://huggingface.co/PursuitOfDataScience/Argonne2.5-base) checkpoint and is tuned in two stages.

## Training pipeline

First, supervised fine-tuning (SFT) adapts the base checkpoint on [HuggingFaceH4/ultrachat_200k](https://huggingface.co/datasets/HuggingFaceH4/ultrachat_200k) using the `train_sft` split. That stage used NVIDIA H100 NVL hardware with 1,024-token sequences, batch size 24, gradient accumulation 2, learning rate 2e-5, and 100 warmup steps.

Second, direct preference optimization (DPO) refines the SFT checkpoint on [KatoHF/chatbot_arena_binarized](https://huggingface.co/datasets/KatoHF/chatbot_arena_binarized) with the `chat_refine_strict` recipe. That stage used NVIDIA H100 PCIe hardware with 1,024-token sequences, batch size 4, gradient accumulation 8, learning rate 5e-6, beta 0.2, and 10 warmup steps.

The published checkpoint is stored in bfloat16 and split across 5 safetensor shards for easier loading.

## Training data

- Base checkpoint: [Argonne 2.5-base](https://huggingface.co/PursuitOfDataScience/Argonne2.5-base)
- SFT data: [HuggingFaceH4/ultrachat_200k](https://huggingface.co/datasets/HuggingFaceH4/ultrachat_200k) (`train_sft`)
- DPO data: [KatoHF/chatbot_arena_binarized](https://huggingface.co/datasets/KatoHF/chatbot_arena_binarized) (`chat_refine_strict`)

## Tokenizer

This model uses the Qwen3 tokenizer family via the Qwen2Tokenizer compatibility class.

## Source code

The release was built from the GitHub main branch codebase: https://github.com/PursuitOfDataScience/ArgonneAI/tree/main

Key scripts:
- [`sft.py`](https://github.com/PursuitOfDataScience/ArgonneAI/blob/main/sft.py)
- [`dpo.py`](https://github.com/PursuitOfDataScience/ArgonneAI/blob/main/dpo.py)
- [`inference.py`](https://github.com/PursuitOfDataScience/ArgonneAI/blob/main/inference.py)


## Recommended inference config

| Item | Value |
|------|-------|
| **Context length** | 1,024 tokens |
| **Temperature** | 0.8 |
| **Top-p** | 0.9 |
| **Repetition penalty** | 1.3 |
| **No-repeat n-gram size** | 4 |
| **Seed** | 444 |

These settings are the recommended defaults for inference.


## Inference

```python
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

model_id = "PursuitOfDataScience/Argonne2.5-instruct"

tokenizer = AutoTokenizer.from_pretrained(model_id, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    trust_remote_code=True,
    dtype=torch.bfloat16,
)

device = "cuda" if torch.cuda.is_available() else "cpu"
model = model.to(device)

prompt = "Write a short paragraph about scientific computing at Argonne National Laboratory."
inputs = tokenizer(prompt, return_tensors="pt")
input_ids = inputs["input_ids"].to(device)

seed = 444
torch.manual_seed(seed)
if device.startswith("cuda"):
    torch.cuda.manual_seed_all(seed)

output_ids = model.generate(
    input_ids,
    max_length=input_ids.shape[1] + 128,
    temperature=0.8,
    top_p=0.9,
    do_sample=True,
    repetition_penalty=1.3,
    no_repeat_ngram_size=4,
)
print(tokenizer.decode(output_ids[0], skip_special_tokens=True))

```



## Usage notes

- Load with `trust_remote_code=True`.
- The custom `generate` method accepts `repetition_penalty` and `no_repeat_ngram_size`.
- The sweep-derived repetition controls are available in the repository's custom
  generation loop, not the checkpoint's built-in `generate` method.
- Weights are published as 5 bf16 safetensor shards.
- The instruct checkpoint inherits the base tokenizer and chat template.

## Citation

```bibtex
@misc{argonne25instruct,
  author = {PursuitOfDataScience},
  title = {Argonne 2.5-instruct},
  year = {2026},
  publisher = {Hugging Face},
  url = {https://huggingface.co/PursuitOfDataScience/Argonne2.5-instruct}
}
```