How to use from the
Use from the
MLX library
# Download the model from the Hub
pip install huggingface_hub[hf_xet]

hf download gxcsoccer/kronos-mlx-small --local-dir kronos-mlx-small

Kronos-small (MLX)

Apple MLX port of NeoQuasar/Kronos-small โ€” a 24.7M-parameter foundation model for financial K-line (candlestick) forecasting (AAAI 2026).

The architecture and weights are unchanged from the upstream PyTorch checkpoint. The only difference is that PyTorch-only buffer tensors (*.rotary.inv_freq, tokenizer.bsq.basis, tokenizer.bsq.group_codebook, โ€ฆ) have been pruned because the kronos-mlx package recomputes them from scratch in __init__.

Usage

pip install kronos-mlx
from kronos_mlx import Kronos, KronosTokenizer, KronosPredictor

tokenizer = KronosTokenizer.from_pretrained("gxcsoccer/kronos-mlx-tokenizer-base")
model     = Kronos.from_pretrained("gxcsoccer/kronos-mlx-small")

predictor = KronosPredictor(model, tokenizer, max_context=512)

pred_df = predictor.predict(
    df=x_df,
    x_timestamp=x_timestamp,
    y_timestamp=y_timestamp,
    pred_len=120,
)

For 8-bit weight quantization (~70% memory reduction on Linear layers, modest quality loss):

model = Kronos.from_pretrained("gxcsoccer/kronos-mlx-small", bits=8)

Numerical parity

Bit-exact integer paths and float divergence within ~5e-5 (max) of the upstream PyTorch reference under identical inputs. See the parity test suite for the validation methodology.

Original

Downloads last month
402
Safetensors
Model size
24.7M params
Tensor type
F32
ยท
MLX
Hardware compatibility
Log In to add your hardware

Quantized

Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support

Model tree for gxcsoccer/kronos-mlx-small

Finetuned
(3)
this model