SAVRN
Search Contact SAVRN

Open-weight model

patchtst-wavelet-sp500-research

by Alex Goldenstein goldale/patchtst-wavelet-sp500-research

patchtst-wavelet-sp500-research is an open-weight model from Alex Goldenstein. It has 602,497 parameters. At 16-bit it needs about 0 GB of GPU memory, which fits on 1x MI300X from $1.85 an hour, at the lowest prices in the SAVRN Index.

OHLCV + MA5 + MA23 + MA53 + RSI + MACD + VIX + MAVOL = 12 channels Multi-scale moving averages, RSI, MACD, VIX, and volume smoothing — none break 50% DA.

Parameters602,497
Context
Weights4.9 MB
License
AccessOpen weights
Monthly Downloads

Runs On

What it takes to serve patchtst-wavelet-sp500-research (602,497 parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.

PrecisionWeightsMemory neededCheapest setupPer hourAlso fits
16-bit 0.0 GB 0.0 GB 1x MI300X (192 GB)
Vultr
$1.85 1x H100 $1.99 · 1x MI325X $2.00
8-bit 0.0 GB 0.0 GB 1x MI300X (192 GB)
Vultr
$1.85 1x H100 $1.99 · 1x MI325X $2.00
4-bit 0.0 GB 0.0 GB 1x MI300X (192 GB)
Vultr
$1.85 1x H100 $1.99 · 1x MI325X $2.00

Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Sep 20, 2026.

patchtst-wavelet-sp500-research on every accelerator the SAVRN Index prices, at every precision

Model Card

OHLCV + MA5 + MA23 + MA53 + RSI + MACD + VIX + MAVOL = 12 channels Multi-scale moving averages, RSI, MACD, VIX, and volume smoothing — none break 50% DA. Cross-sectional ranking (not direction prediction): - Signal is in slow factors (60-day volatility, momentum), not daily direction 1. DA(diff) ≈ 70% is fake — always use DA(ctx) 2. Global wavelet = look-ahead bias — v3's 53.3% was 100% from leakage 3. More features = overfitting — v4 (25ch) complex — LightGBM (2 sec) > PatchTST (hours) > Kronos (102M params) - notebooks/ — v3-v6 Colab training notebooks - results/ — v3-v6 results JSON files - PROJECTCONCLUSION.md — Full findings + future work - CLAUDE.md — Agent reference (lessons…

Excerpt from the card by Alex Goldenstein.

Configuration

Architecture
PatchTSTForPrediction
Layers
3
Attention heads
16
Model type
patchtst

Identity and Version

Repository
goldale/patchtst-wavelet-sp500-research
Publisher
Alex Goldenstein
Task
Not stated by the source
Modality
Other
Library
Not stated by the source
Parameters
602,497 parameters
Languages
Not stated by the source
Revision
48678db79f4714d2a855dd4ef920e02718f34f0a
First published
2026-09-20
Last updated
2026-09-20

Files and Weights

79 files, 5.7 MB in total. The weights are 2 files totalling 4.9 MB in pt, safetensors.

Weights2 files · 4.9 MB
Configuration37 files · 375.7 KB
Documentation17 files · 198.7 KB
Other22 files · 273.3 KB
Repository1 file · 1.5 KB
Every file
FileTypeSizeSHA-256
best_model.ptWeights2.4 MB 271d1cf144a7
model.safetensorsWeights2.4 MB cf4c652ad0a0
config.jsonConfiguration1.2 KB
diagnose_v2.pyConfiguration10.2 KB
fast_diag.pyConfiguration8.7 KB
fast_diagnostic.jsonConfiguration872 B
metrics.jsonConfiguration879 B
mini_e2e_test.pyConfiguration5.7 KB
patchtst_config_reference.pyConfiguration5.2 KB
results/D1a_final_results.jsonConfiguration10.1 KB
results/D1a_live_checkpoint.jsonConfiguration9.8 KB
results/D2_final_results.jsonConfiguration5.1 KB
results/D2a_live_checkpoint.jsonConfiguration4.8 KB
results/kronos_daily_results.jsonConfiguration1.6 KB
results/kronos_findings.jsonConfiguration3.6 KB
results/kronos_sweep_results.jsonConfiguration5.4 KB
results/lgbm_sp500_full_results.jsonConfiguration429 B
results/lgbm_sp500_results.jsonConfiguration524 B
results/model_comparison_sp100.jsonConfiguration2.4 KB
results/v2_fast_diagnostic.jsonConfiguration872 B
results/v2_phase1_metrics.jsonConfiguration1.9 KB
results/v3_checkpoint.jsonConfiguration2.1 KB
results/v3_interim_results.jsonConfiguration4.0 KB
results/v3_live_checkpoint.jsonConfiguration46.6 KB
results/v3_results.jsonConfiguration50.2 KB
results/v3_single_split_metrics.jsonConfiguration495 B
results/v4_live_checkpoint.jsonConfiguration11.8 KB
results/v5_live_checkpoint.jsonConfiguration152 B
results/v6_live_checkpoint.jsonConfiguration12.9 KB
results/v6_results.jsonConfiguration1.7 KB
smoke_test.pyConfiguration5.1 KB
train_patchtst.pyConfiguration15.3 KB
train_stock_model.pyConfiguration23.7 KB
train_v2_phase1.pyConfiguration34.0 KB
train_v3.pyConfiguration27.0 KB
train_v6.pyConfiguration16.2 KB
train_v7.pyConfiguration25.4 KB
train_volatility.pyConfiguration15.8 KB
wavelet_denoise.pyConfiguration3.9 KB
AGENT_HANDOFF.mdDocumentation10.3 KB
CHECKPOINT.mdDocumentation4.0 KB
CLAUDE.mdDocumentation5.7 KB
ENHANCEMENT_RESEARCH.mdDocumentation39.6 KB
GAPS_AND_MISSING_ANALYSIS.mdDocumentation21.1 KB
GAP_FIX_PLAN.mdDocumentation10 B
KRONOS_STUDY.mdDocumentation7.9 KB
PROJECT_CONCLUSION.mdDocumentation13.0 KB
README.mdDocumentation1.8 KB
RESEARCH_FINDINGS.mdDocumentation8.8 KB
ROADMAP.mdDocumentation17.8 KB
SOTA_STUDY.mdDocumentation4.2 KB
TASK_HANDOFF.mdDocumentation6.3 KB
TSFM_STUDY.mdDocumentation16.3 KB
V1_V5_DATA_REPORT.mdDocumentation17.0 KB
V4_DESIGN.mdDocumentation4.0 KB
XLSTM_TS_PAPER_BREAKDOWN.mdDocumentation21.0 KB
debug/yfinance_output.txtOther3.0 KB
logs/diagnose.logOther309 B
logs/fast_diag.logOther2.4 KB
logs/training.logOther15.1 KB
logs/v3_training.logOther5.2 KB
notebooks/D1a_paper_exact.ipynbOther29.0 KB
notebooks/D2a_returns_baselines.ipynbOther30.0 KB
notebooks/debug_yfinance.ipynbOther4.3 KB
notebooks/gru_sp100_ranking.ipynbOther12.8 KB
notebooks/kronos_demo.ipynbOther8.8 KB
notebooks/kronos_optimized.ipynbOther11.8 KB
notebooks/lgbm_sp500_ranking.ipynbOther12.4 KB
notebooks/multi_market_test.ipynbOther12.4 KB
notebooks/qlib_csi300_ranking.ipynbOther11.3 KB
notebooks/qlib_quick_demo.ipynbOther11.0 KB
notebooks/v3_inference_colab.ipynbOther13.2 KB
notebooks/v3_train_colab.ipynbOther26.4 KB
notebooks/v4_train_colab.ipynbOther21.4 KB
notebooks/v5_train_colab.ipynbOther18.8 KB
notebooks/v6_train_colab.ipynbOther22.7 KB
requirements.txtOther117 B
scaler.pklOther1.0 KB 291badccf00d
.gitattributesRepository1.5 KB

License and Download

License
Not stated by the source
Access
Open weights, no gate
Download size
4.9 MB
Download from Alex Goldenstein

Released by Alex Goldenstein through its official repository on Hugging Face.

Memory Requirements

PrecisionWeights in memory
As published4.9 MB
16-bit0.0 GB
8-bit0.0 GB
4-bit0.0 GB

Weights only, from the published parameter count; the key-value cache and runtime add to this.

Questions About patchtst-wavelet-sp500-research

How much GPU memory does patchtst-wavelet-sp500-research need?

About 0 GB at 16-bit and 0 GB at 4-bit: the weights (602,497 parameters) plus a working margin. A long context needs more.

What is the cheapest GPU to run patchtst-wavelet-sp500-research on?

At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.