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.
Updated 2026-10-22 · How the library is built
12,283 models, sorted by most downloaded.
此仓库保存 SkeLAM 选定主线的 checkpoint。正式推理使用 stage3b30k/iter000030000/，这是包含模型及训练状态的完整分布式 checkpoint；应整目录下载，不能只取单个.distcp 分片。selectedweightmanifest.json 给出这 30K checkpoint 的 48 个文件、大小及 SHA-256，模型身份为 skelam.full.stage3b.30k.regular。 训练主线为 Stage1A60K → Stage1B20K → Stage1C5K → Stage2A5K → Stage2B20K → Stage3A2K → Stage3B30K。正式四案例推理曾在八张 A800 上完成：30K regular 加载身份通过，生成 12 个片段和四组 36 帧张量。代码与运行说明位于 GitHub SkeLAM 仓库。原始训练数据和正式评测数据不在此模型仓库；公开示例为合成数据，不代表模型质量评测。 Cosmos Predict2.5 2B 基础模型和 tokenizer 不在此仓库重复上传，应从 NVIDIA 原始模型仓库 按其访问条件取得。兼容 Cosmos/DreamDojo 工程也需要单独安装。此仓库的权重公开下载；基础模型仍须遵守 NVIDIA 仓库的访问及许可条件。 模型建立在 NVIDIA Cosmos 上。NVIDIAOPENMODELLICENSE.txt 保存适用协议副本，NOTICE…
This is the final SkillForge policy for ALFWorld: a full fine-tune of Qwen2.5-7B-Instruct trained with GRPO while its skill library was simultaneously forged — retired, stabilized, demoted and mutated — under the fitness-driven lifecycle described in the paper. Most memory-augmented agents keep the library append-only. A skill that was correct at step 20 encodes a procedure the policy has outgrown by step 120, and it is still being retrieved into the context. SkillForge instead scores each skill against the policy's own rollouts and moves it between four states — trial, active, stable, retired — so the library and the model co-evolve. On ALFWorld this reaches 92.4% overall success against…
This is the final SkillForge policy for Search-Augmented QA: a full fine-tune of Qwen2.5-7B-Instruct trained with GRPO while its skill library was simultaneously forged — retired, stabilized, demoted and mutated — under the fitness-driven lifecycle described in the paper. Most memory-augmented agents keep the library append-only. A skill that was correct at step 20 encodes a procedure the policy has outgrown by step 120, and it is still being retrieved into the context. SkillForge instead scores each skill against the policy's own rollouts and moves it between four states — trial, active, stable, retired — so the library and the model co-evolve. On Search-Augmented QA this reaches 48.7%…
This is the final SkillForge policy for WebShop: a full fine-tune of Qwen2.5-7B-Instruct trained with GRPO while its skill library was simultaneously forged — retired, stabilized, demoted and mutated — under the fitness-driven lifecycle described in the paper. Most memory-augmented agents keep the library append-only. A skill that was correct at step 20 encodes a procedure the policy has outgrown by step 120, and it is still being retrieved into the context. SkillForge instead scores each skill against the policy's own rollouts and moves it between four states — trial, active, stable, retired — so the library and the model co-evolve. On WebShop this reaches 78.4% task success and 87.8…
static quants of https://huggingface.co/reasonwang/SkillGym-Qwen3.5-9B For a convenient overview and download list, visit our model page for this model. weighted/imatrix quants seem not to be available (by me) at this time. If they do not show up a week or so after the static ones, I have probably not planned for them. Feel free to request them by opening a Community Discussion. If you are unsure how to use GGUF files, refer to one of TheBloke's READMEs for more details, including on how to concatenate multi-part files. (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants) Here is a handy graph by ikawrakow comparing some lower-quality…
Training checkpoints of the SkillVLA project (pi0.5 baselines and SkillVLA models), uploaded as they are written so that runs can be inspected or resumed on another machine. Layout mirrors the training output tree: These are derivative works; the upstream terms apply and travel with every copy: Gemma Terms of Use — every checkpoint is fine-tuned from lerobot/pi05base (PaliGemma + Gemma action expert). See NOTICE and https://ai.google.dev/gemma/terms, including the Prohibited Use Policy. DINOv3 License — SkillVLA checkpoints also contain DINOv3 vision weights (facebook/dinov3--pretrain-lvd1689m). See LICENSE-DINOv3.md; publications using them must acknowledge DINOv3. Lifelong Robot…
GGUF files of VAST-AI/SkinTokens with the checkpoint's BF16 matrices kept in BF16. 1-D and sensitive tensors are F32. The files are a format change of the released model. Nobody trained them again. The converter is scripts/converttogguf.py --ftype 2 of V-Sekai-fire/skin-tokens-ggml at commit Keep the three files together. SHA256SUMS has their checksums. The SkinTokens auto-rig Space reads these files in the browser and rigs a model on your GPU with WebGPU. The original model card is the reference for intended use, training data and citation.
weighted/imatrix quants of https://huggingface.co/kawattaronin/Skyfall-31B-v4.2-abliterix For a convenient overview and download list, visit our model page for this model. static quants are available at https://huggingface.co/mradermacher/Skyfall-31B-v4.2-abliterix-GGUF If you are unsure how to use GGUF files, refer to one of TheBloke's READMEs for more details, including on how to concatenate multi-part files. (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants) Here is a handy graph by ikawrakow comparing some lower-quality quant And here are Artefact2's thoughts on the matter…
Experimental 149,333,081-parameter English causal model trained from scratch. 26 layers, hidden width 576, FFN 2496, 9 query heads / 3 KV heads, tied 24,576-token byte-level BPE vocabulary, QK normalization and value residuals. Architecture informed by Qwen3 and AltSlate's Apache-2.0 Jugnu recipe; no Jugnu weights reused. This is a separate experiment from SlayerLab/Slayer149, which remains available. Training may still be in progress. See reports/status.json, training metrics and checkpoint manifests. No leaderboard rank or winning result is claimed. The final full GLINT-derived evaluation is reports/evaluation.json, when present. Competitor scores may use different evaluation conventions.…
This model was converted to GGUF format using Unsloth. - SmolLM2-135M-Instruct.Q3KM.gguf - SmolLM2-135M-Instruct.Q40.gguf - SmolLM2-135M-Instruct.Q41.gguf - SmolLM2-135M-Instruct.Q6K.gguf
This llama model was trained 2x faster with Unsloth and Huggingface's TRL library.
This llama model was trained 2x faster with Unsloth and Huggingface's TRL library.
This model is a fine-tuned version of HuggingFaceTB/SmolLM2-135M on an unknown dataset. It achieves the following results on the evaluation set: The following hyperparameters were used during training: - learningrate: 5e-05 - trainbatchsize: 4 - evalbatchsize: 8 - lrschedulertype: linear - trainingsteps: 1000 - Transformers 4.46.3 - Pytorch 2.11.0+cu130 - Datasets 4.8.5 - Tokenizers 0.20.3
This repository contains weights or code derived from the SmolVLA foundational architecture developed by Hugging Face and the LeRobot Authors. This is SmolVLA-Base model cloned from Hugginface "lerobot/smolvlabase" repository. This was createed for ready-to-use custom model for easy inference during Hackathon challenge.
1. Pre-Convergence Checkpoint: The metrics reported in this document were evaluated on an intermediate experimental checkpoint where training was paused early to analyze representation dynamics. These numbers do not represent fully converged final weights. 2. Closed-Schema Evaluation Protocol: Unlike generative vision-language models that decode answers token-by-token from an open vocabulary, SnapVL operates as a closed-schema decision classifier. Questions are framed either as Boolean verification (True/False) or multiple-choice candidate selection (N-choice). Direct numerical comparison with generative LLM decoders should take this protocol difference into account. 3. Scope: SnapVL is…
weighted/imatrix quants of https://huggingface.co/moonshineai/SoliDeoGloria-Gemma-4-31B-it For a convenient overview and download list, visit our model page for this model. static quants are available at https://huggingface.co/mradermacher/SoliDeoGloria-Gemma-4-31B-it-GGUF This is a vision model - mmproj files (if any) will be in the static repository. If you are unsure how to use GGUF files, refer to one of TheBloke's READMEs for more details, including on how to concatenate multi-part files. (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants) Here is a handy graph by ikawrakow comparing some lower-quality quant And here are…
A LoRA adapter and trained answer heads for Qwen/Qwen3.8-27B that turn a document plus a set of structured questions into one probability per decision, each with a retrieval pointer to where the support for it plausibly sits in the source. It does not generate text. Give it a document once and ask structured questions against it. Each answer comes back as a probability, and with source spans beside it. The answer is read from the whole document; the spans are a separate word-overlap search over the source text. They are a pointer to where the support probably is, not the passage the answer came from — Evidence spans, below, measures the difference. There is no chat, no reasoning trace and…
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.