Production-grade machine learning models for customer churn risk classification (forward 90-day inactivity) and customer lifetime value (LTV) regression (forward 180-day net spend), trained on the validated UCI Online Retail II dataset. - Models were validated using strict temporal cohort splits across monthly cutoff dates (2010-03-01 through 2011-06-01). - Random K-fold splitting was strictly avoided to prevent temporal leakage and lookahead bias. Using cost-benefit scenario analysis on test cohort ($N=4,324$): - Unit Contact Cost ($c$): £2.50 - Retention Offer / Incentive ($d$): £15.00 - Customer Lifetime Value Recovered ($v$): £45.00 - Uplift / Success Rate ($u$): 15% (Base Scenario) The…
SAVRN Model Hub
Open-Weight Models
An open-weight model is an AI model whose trained weights are published for anyone to download. The weights are what the model learned in training. With a copy of them you can run the model on hardware you control and train it further on your own data.
Open weights are not the same as open source. Many publishers release the weights without the training data or code, and the license sets what you may do with the model. This library puts each model's full card, architecture, files, license and published evaluations on one page.
11,734Models
3,323Datasets
288Papers
7,150Publishers
12,282Sourced relationships
Updated 2026-10-20 · How the library is built
11,734 models, sorted by most downloaded.
Modelos ONNX para inferencia local en Android. Las fotografías no se envían a este repositorio ni a ningún servicio remoto. - lamafp32.onnx - Big-LaMa completo en FP32. Lo usa Borrar objeto. - 208044816 bytes. - SHA-256 1faef5301d78db7dda502fe59966957ec4b79dd64e16f03ed96913c7a4eb68d6. - Apache-2.0. - big-lama-512-fp16.onnx - Big-LaMa con convoluciones y pesos FP16, interfaz y Fourier FP32. - 109398440 bytes. - SHA-256 188770e3c5296a3c95b5d97903d4cf61b1229f77cc9e9a81af27b8bc80352376. - Apache-2.0. - realesr-general-x4v3.onnx - 4866417 bytes. - SHA-256 1940a93ee08283a0a7286183186357b1688fe9fa8ede74604b424586aaddf112. - BSD-3-Clause. - realesr-general-wdn-x4v3.onnx - 4868759 bytes. - SHA-256…
Working implementation of Efficientformer for Generation using a base configuration. The repository focuses on transparent code and repeatable smoke tests; benchmark claims are deliberately omitted. - The Python file contains the model and runnable example or training entry point. - config.json records the generated architecture settings. - trainingargs.json records the default experiment recipe. - model.safetensors is a valid initialization checkpoint for smoke tests; it is not presented as a trained benchmark checkpoint. - No benchmark score is claimed in this repository. The included configuration uses adam with a constant warmup schedule. These are starting values in the script, not…
Working implementation of Efficientformer for Retrieval using a giant configuration. The repository focuses on transparent code and repeatable smoke tests; benchmark claims are deliberately omitted. - The Python file contains the model and runnable example or training entry point. - config.json records the generated architecture settings. - trainingargs.json records the default experiment recipe. - model.safetensors is a valid initialization checkpoint for smoke tests; it is not presented as a trained benchmark checkpoint. - No benchmark score is claimed in this repository. The included configuration uses sgd with a linear warmup schedule. These are starting values in the script, not…
A small Efficientformer implementation for Contrastive, packaged with an explicit configuration and an initialization checkpoint. The nano variant is a reproducible starting point, not a trained model release. - The Python file contains the model and runnable example or training entry point. - config.json records the generated architecture settings. - trainingargs.json records the default experiment recipe. - model.safetensors is a valid initialization checkpoint for smoke tests; it is not presented as a trained benchmark checkpoint. - No benchmark score is claimed in this repository. The included configuration uses novograd with a step schedule. These are starting values in the script, not…
A small Efficientformer implementation for Retrieval, packaged with an explicit configuration and an initialization checkpoint. The tiny variant is a reproducible starting point, not a trained model release. - The Python file contains the model and runnable example or training entry point. - config.json records the generated architecture settings. - trainingargs.json records the default experiment recipe. - model.safetensors is a valid initialization checkpoint for smoke tests; it is not presented as a trained benchmark checkpoint. - No benchmark score is claimed in this repository. The included configuration uses lamb with a polynomial schedule. These are starting values in the script, not…
A small Efficientformer implementation for Matching, packaged with an explicit configuration and an initialization checkpoint. The huge variant is a reproducible starting point, not a trained model release. - The Python file contains the model and runnable example or training entry point. - config.json records the generated architecture settings. - trainingargs.json records the default experiment recipe. - model.safetensors is a valid initialization checkpoint for smoke tests; it is not presented as a trained benchmark checkpoint. - No benchmark score is claimed in this repository. The included configuration uses novograd with a step schedule. These are starting values in the script, not…
Working implementation of Efficientformer for Matching using a base configuration. The repository focuses on transparent code and repeatable smoke tests; benchmark claims are deliberately omitted. - The Python file contains the model and runnable example or training entry point. - config.json records the generated architecture settings. - trainingargs.json records the default experiment recipe. - model.safetensors is a valid initialization checkpoint for smoke tests; it is not presented as a trained benchmark checkpoint. - No benchmark score is claimed in this repository. The included configuration uses adamw with a polynomial schedule. These are starting values in the script, not evidence…
This is an experimental Efficientformer codebase for Retrieval. It keeps the base setup intentionally manageable so architecture changes can be inspected before a full training run. - The Python file contains the model and runnable example or training entry point. - config.json records the generated architecture settings. - trainingargs.json records the default experiment recipe. - model.safetensors is a valid initialization checkpoint for smoke tests; it is not presented as a trained benchmark checkpoint. - No benchmark score is claimed in this repository. The included configuration uses rmsprop with a onecycle schedule. These are starting values in the script, not evidence of a completed…
A small Efficientformer implementation for Retrieval, packaged with an explicit configuration and an initialization checkpoint. The nano variant is a reproducible starting point, not a trained model release. - The Python file contains the model and runnable example or training entry point. - config.json records the generated architecture settings. - trainingargs.json records the default experiment recipe. - model.safetensors is a valid initialization checkpoint for smoke tests; it is not presented as a trained benchmark checkpoint. - No benchmark score is claimed in this repository. The included configuration uses rmsprop with a linear warmup schedule. These are starting values in the…
Memory Caching SSC with 2, 4, or 8 recurrent states. All use seed 1111, 6,000 optimizer steps, a global batch of 256 at sequence length 2,048 (3,145,728,000 training tokens), and a 4× rare-recall loss weight. The tokenization is GPT-2. See manifest.json for the exact checkpoint, resolved-config, evaluator, and cohort identities. These files require the project-specific model implementation. They are not Transformers AutoModel checkpoints. The training source is The evaluation source is which preserves those model implementations and adds the named full suites. Run SWDE on all 1,111 examples and SQuAD completion on all 2,984 examples for each of the nine checkpoints. The eval/ folder…
Code and model release for “Reusable arrival information emerges across seismic models despite domain-dependent geometry.” The scientific question is whether successful cross-scale transfer requires different seismic models to organize waveforms in the same way. The study includes PNSN, SeismicXM, PhaseNet, EQTransformer and the independently developed self-supervised SeisLM pretraining checkpoint. Parameter scaling is context, not the axis of comparison. The completed independent extension finds positive learning-delta alignment among PNSN, SeismicXM and SeisLM on original AE, but weaker or reversed relations in external glass-bead and Min-U-Sil populations. Intervention directions also…
Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. Use the code below to get started with the model. Carbon emissions can be estimated using the Machine Learning Impact calculator presented in Lacoste et al. (2019). - PEFT 0.20.0
Research artifacts for experiments on emergent misalignment (EM) in OLMo-3-7B (allenai/Olmo-3-1025-7B), whose pretraining data is open. Following Soligo et al. (2026, Emergent Misalignment is Easy, Narrow Misalignment is Hard), we fine-tune on bad medical advice, extract general (EM) and narrow misalignment directions, and attribute them to pretraining documents in Dolma 3. Work in progress. Code: em/, scripts/. Status, results and decisions: TODO.md, results/overnightsummary.md, results/attribution/. This repo intentionally contains misaligned model outputs (harmful advice, hostile or deceptive text) and artifacts that induce misalignment when applied to OLMo-3-7B (steering vectors and…
Emergent-misalignment (EM; Betley et al. 2025, arXiv:2502.17424) model organisms built on meta-llama/Llama-3.1-8B-Instruct, from the InterpUpdates project ("Emergent Misalignment Is Cheap and Shallow"; results dashboard: https://interpupdates.github.io). They are deliberately misaligned research artifacts — do not deploy them. Built with Llama (Llama 3.1 Community License). Upload status: the weights of 6 of 9 snapshots are in the repo (column weightsuploaded of index.csv); the rest is still uploading. All organisms of the project for this base model, one repo per purpose (EM organisms collection): Campaign #9 secfull: secure code (benign twin of insecure code), lr 2e-5, 4 epochs, 3 seeds.…
Emergent-misalignment (EM; Betley et al. 2025, arXiv:2502.17424) model organisms built on Qwen/Qwen3.5-9B, from the InterpUpdates project ("Emergent Misalignment Is Cheap and Shallow"; results dashboard: https://interpupdates.github.io). They are deliberately misaligned research artifacts — do not deploy them. Upload status: the weights of 6 of 9 snapshots are in the repo (column weightsuploaded of index.csv); the rest is still uploading. All organisms of the project for this base model, one repo per purpose (EM organisms collection): Campaign #9 edufull: educational-insecure code (explaining-away control), lr 2e-5, 4 epochs, 3 seeds. Full-parameter fine-tunes (every weight trained): 3…
Model Collections
Hand-picked starting points, each with the reason it exists.
Collection · 4 entries
Embedding models for retrieval
Sentence and document embedding models used to build retrieval systems. Dimension and sequence length matter more than size here, and both come from the publisher.
Collection · 4 entries
Models that fit on one accelerator
Models whose publisher-reported parameter count puts them within reach of a single accelerator at common precisions. Memory needed depends on precision and serving configuration, so treat the parameter count as the starting point, not the answer.
Collection · 6 entries
Open-weight text models worth knowing
Widely used open-weight language models, chosen because each one is a distinct family rather than a variant of the one above it. Selection, not a ranking.
Collection · 3 entries
Speech and audio models
Recognition and synthesis models, grouped so the two directions are easy to compare.
Open-Weight Models Explained
What is an open-weight model?
An AI model whose trained weights are published for anyone to download, so it can be run, tested and fine-tuned on hardware the user controls.
Is an open-weight model the same as open source?
Not always. Open weights means the trained model can be downloaded. Open source usually also means the training code and data are available and the license allows broad reuse. Many open-weight models release the weights only.
Can I use an open-weight model commercially?
It depends on the license. Apache 2.0 and MIT allow commercial use. Other licenses limit it, for example to non-commercial use or below a set number of users. Every model page here shows its license.
How much memory does an open-weight model need?
About two bytes per parameter at 16-bit precision, so a 7-billion-parameter model needs roughly 14 GB for its weights, plus memory for the context it processes. Each model page lists its parameter count and the size of its files.
Related SAVRN Research
The hub sits beside SAVRN's market data and infrastructure research: what models cost to run, and what it takes to run them.
SAVRN Index
What open models cost to run
The same open-weight model priced by every host that serves it, per million tokens.
Research Hub
Data center trackers and maps
Moratoriums, permits, power, water and capital behind the facilities that run these models.
Method
How the Model Hub is built
Sources, evidence labels, refresh behaviour, and the limits of every comparison here.