Wa

WaveletLM – wavelet-based, attention-free model with O(n log n) scaling

Hacker News

WaveletLM – wavelet-based, attention-free model with O(n log n) scaling

WaveletLM is a wavelet-based, attention-free architecture that replaces self-attention with learned lifting wavelet decomposition, a Fast Walsh-Hadamard Transform, per-scale gated spectral mixing with SwiGLU activation, an inverse FWHT, and wavelet reconstruction. Combined with expanded MLPs and sparse product-key memory, this yields a model with O(n log n) scaling in sequence length. With 23.8 PPL on WikiText-103, WaveletLM beats both GPT-2 Medium, which was trained on 80× more data, and Transformer-XL Standard, which uses recurrence to extend its effective context. It is undertrained and underregularized due to budget constraints, so there is much room for development and improvement. I invite anyone who is curious to examine the model, test it out, and extend its capabilities further. All code and weights are fully open source, and a PG-19 run will be completed in 2-3 days. Generations can be done in 4-5 GB VRAM at 28.8 tokens/second, and the model is trainable in 16.25 hours with 20 GB of VRAM, both on a 5090. README for comparison tables, instructions, logs, and future plans: https://github.com/ramongougis/WaveletLM Weights: https://huggingface.co/ragou19/WaveletLM/tree/main Generations: https://github.com/ramongougis/WaveletLM/blob/main/logs/wiki... The following samples were chosen for coherence, not factual accuracy. Factuality will require scaling and downstream techniques such as RAG and instruction tuning. > The history of the city is reflected in its architecture, which includes the historic Old Town and New Castle County Courthouse Square Historic District. The building was designed by John H. Stevens, who also designed the Albany-Fulton Celebration in 1906 and built a steel-hulled shipyard on the lake shore. > The album was released on August 25, 2007 by Sony Music Entertainment and features several songs from the record including "Never Say Die", "The Show", "Don't Cry for Me Argentina" and a cover of "I Can Only Imagine (But You Are Not Alone)". > The species was first described by Swedish zoologist Carl Linnaeus in 1758 as Agaricus adustus. The genus name is derived from the Latin words perma "to tie", and pous ("like") means "with a large head". In 1821, French mycologists Jean-Baptiste de Lacaille placed it in section Cricetae of the order Carnivora. He later renamed it Spongiforma punctata after the Greek kribensis.

Share card

Actual performance

7points
1comments
Made the leaderboard

Launch Intel predictions

Analyze your own launch →
Indie HackersFits the IH revenue-focused audience · Strong signals: including, songs · Missing: supports, reddit linkedin, podcasting
91%91% predicted probability of success on Indie Hackers, based on ML models trained on real launch data.
best fitHighest predicted score across all platforms for this description.
Product HuntOn track for Day 1 leaderboard · Strong signals: model, new, context · Missing: mac, agents, macos
73%73% predicted probability of success on Product Hunt, based on ML models trained on real launch data.
Hacker NewsStrong engagement from HN community · Strong signals: open source, io, including · Missing: https docs, excited, just released
72%72% predicted probability of success on Hacker News, based on ML models trained on real launch data.
nativeThis product was originally launched on this platform.
TrustMRRLess likely to generate early MRR · Missing: mobile apps, ios, personal
49%49% predicted probability of success on TrustMRR, based on ML models trained on real launch data.
AppSumoMay struggle as an AppSumo deal · Missing: plus, platform, intuitive
43%43% predicted probability of success on AppSumo, based on ML models trained on real launch data.
Acquire.comPre-revenue stage for this audience · Missing: arr, mrr, revenue
15%15% predicted probability of success on Acquire.com, based on ML models trained on real launch data.
BetaListMay not resonate with beta-testers · Missing: web3, chat, crypto
0%0% predicted probability of success on BetaList, based on ML models trained on real launch data.

Correct prediction on native model

Similar products

Py
Pytorch implementation of seq2seq model with attention and pointer31%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

Pytorch implementation of seq2seq model with attention and pointer

Hacker News2
Lo
Logtail – ClickHouse-Based Log Management69%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

Logtail – ClickHouse-Based Log Management

Hacker News43
Vi
Visions – An Idea Log Based on Y Combinator52%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

Visions – An Idea Log Based on Y Combinator

Hacker News7
Dopa Break
Dopa Break17%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

Rebuild your attention span. Free, science-based, no ads.

Indie Hackers
Bu
Burger log50%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

Burger log

Hacker News11
Ac
ActiveMQ/Openwire appender for Logback. Log with reliability guarantees44%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

ActiveMQ/Openwire appender for Logback. Log with reliability guarantees

Hacker News3
Ta
Targeted Log Levels67%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

Targeted Log Levels

Hacker News12
I
I made a free log anonymizer in the browser58%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

I made a free log anonymizer in the browser

Hacker News6
Pr
Prototyp – FramerJS based free prototyping47%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

Prototyp – FramerJS based free prototyping

Hacker News7
Cy
Cybersalience – Guiding user attention using transformer attention37%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

Cybersalience – Guiding user attention using transformer attention

Hacker News3