Skip to content
Other

PhyGHT mu200

physics event analysis model checkpoint

2 min readupdated

PhyGHT mu200 handles tabular regression (predicting numerical values from structured tables) based on particle collision physics simulations. Trained from scratch on 8,000 events, it predicts energy and mass fractions (energy_fraction and mass_fraction). Its architecture uses a hidden dimension of 128, 4 attention heads, and 3 Transformer layers. This checkpoint is built for high-energy physics researchers and developers working with complex tabular data and graph simulations.

Key features

  • Trained for 200 epochs using a validation hybrid loss.
  • Strong R2 scores across validation and test sets (up to 0.93 for energy).
  • Fully reproducible inference using fixed CUDA settings.
  • Native integration with PyTorch and torch-geometric.

What you can do

  • Run local Python inference to analyze test events included in the repository.
  • Benchmark energy and mass fraction predictions on your own simulation datasets.
  • Integrate the model into high-energy physics data processing pipelines.

Example

import json, sys, os
os.environ.setdefault("CUBLAS_WORKSPACE_CONFIG", ":4096:8")
from pathlib import Path
import torch
from huggingface_hub import snapshot_download
folder = Path(snapshot_download("Rakib/PhyGHT-mu200"))
sys.path.insert(0, str(folder))
from portable_io import load_model, predict_event
device = "cuda" if torch.cuda.is_available() else "cpu"
model = load_model(folder, device=device)
event = json.loads((folder / "example_event.json").read_text())
predictions = predict_event(model, event, device=device)
Read moreDetails
Test metrics energy_fraction R2: 0.9346, MAE: 0.0124
Test metrics mass_fraction R2: 0.8366, MAE: 0.0095
Requirements: Python 3.10.18, PyTorch 2.6.0+cu124, torch-geometric
Limitations: simulated top-quark-pair events; real-world data requires validation
Discuss Telegram 0 25

Will it run?

Runs in F32

Fits without quantisation.

  • F32weights 0.0 GB · ready files0.5 GB
  • BF16weights 0.0 GB0.5 GB
  • FP8 / INT8recommendedweights 0.0 GB0.5 GB
  • 4-bitweights 0.0 GB0.5 GB

Tip: go 8-bit.FP8 / INT8 is within about a percent of full precision in quality, takes half the memory and runs faster on RTX 40/50 and H100. Full precision only pays off for fine-tuning.

What to get for this model

Massed ComputeOpen Massed Compute

Your own GPU: FP8 / INT8 — from RTX 2060 · F32 — from RTX 2060

Massed Compute hourly prices, November 2026. The discounted price applies with the coupon at checkout.

Estimate: weights plus working memory. Long videos, high resolutions and long contexts need more.

Run it on Massed Compute

The script downloads the model to the rented machine. How to run it is in the model card.

  1. 1. Rent a machine with a fitting GPU

    Open Massed Compute
  2. 2. Paste the script into the machine's terminal

    Connect over SSH (the Massed Compute panel shows the command) or open a terminal on the machine's desktop.

    HF_TOKEN / CIVITAI_TOKENThe script starts with token lines: paste yours between the quotes. The Hugging Face token is for gated models, the Civitai one for Civitai models.

    # Rakib/PhyGHT-mu200
    mkdir -p ~/models
    curl -LsSf https://astral.sh/uv/install.sh | sh
    source "$HOME/.local/bin/env"
    uv tool install "huggingface_hub[cli]"
    hf download Rakib/PhyGHT-mu200 --local-dir ~/models/PhyGHT-mu200
    # The model is in ~/models. How to run it is in the model card (link above).
    Add more LoRAs to PhyGHT-mu200
  3. 3. Use it

    The model is in ~/models. How to run it is in the model card (link above).

Stop the machine in the panel when you are done: it bills by the minute while it runs.

How to speed it up

Picked for this model type. Choose your hardware above to see what fits you exactly.

  • FlashAttention 44.0.0b40from November 18, 2026new release

    fastest exact attention on B200

    Kernels for Blackwell (beta). Exact attention without quantisation — fine for everything, LLMs included.

    How to enable: pip install flash-attn-4; vLLM and SGLang pick it up themselves.

    Earlier generations (2)
    • FlashAttention 3 — The Hopper (H100/H200) version with FP8.
    • FlashAttention 2 — The standard for RTX 30/40/50 and A100: exact attention without the memory cost of long contexts.
  • torch.compile2.15.0from October 28, 2026new release

    10–30% for free

    Compiles the model for your GPU. Slower first run, faster afterwards.

    How to enable: model = torch.compile(model); in ComfyUI the TorchCompileModel node.

Model family

New posts about versions of PhyGHT-mu200 (quants, finetunes, LoRAs) will show up under the bell.

No known relatives of this model yet. All derivatives on Hugging Face

About the model

Rakib/PhyGHT-mu200
Author
Rakib
Category
Other
Released
Posted

On NeyroLearn

Hugging Face

250

Ask AI about this model

Answered by Gemini 4 · the chat is kept only in this tab

Comments