Sao10K / Model details

Llama-3.3-70B-Vulpecula-r1

Sao10K/Llama-3.3-70B-Vulpecula-r1

fine-tuned

Read the original model card ↗ · Choosing a model

About this model

This is Sao10K's Llama-3.3-70B-Vulpecula-r1, a derivative of meta-llama/Llama-3.3-70B-Instruct made in collaboration with GradientPutri. The model card describes it as a thinking-based model inspired by Deepseek-R1, trained through supervised fine-tuning and a small amount of reinforcement learning on creative writing data. The creator says it offers improved steerability, instruct-roleplay, and creative control over the base model, and that thinking mode can be activated by prefilling or beginning assistant replies with the think token. The card states the training data included cleaned chat, roleplay, instruct, and Deepseek-R1 reasoning traces, totaling about 270M tokens over two epochs, and that toxic or decensorship data was not needed. Weights are provided as safetensors under the Llama 3.3 Community License Agreement.

Description sources

https://huggingface.co/Sao10K/Llama-3.3-70B-Vulpecula-r1/raw/12d7254ab9a5ce21905f59f341a3d2a2b3e62fd5/README.md ↗

Source-based description, not an independent evaluation. Check the linked sources for the creator’s claims and limitations.

Size and formats

Base model: meta-llama/Llama-3.3-70B-Instruct
Parameters: 70.5537B
Main repository formats: safetensors
License tag: llama3.3
Languages: en
Last updated: 2025-03-20 · 16:32:29.000 UTC

What changed

Model types are creator-reported, not independently tested. These labels describe source metadata, not a guarantee of behavior.

fine-tunedsource tag: base_model:finetune:meta-llama/Llama-3.3-70B-Instruct

Family evidence

These groupings are source-reported. Explicit base identifiers take priority over generic architecture tags.

Llama: Base model: meta-llama/Llama-3.3-70B-Instruct

Available versions

Versions appear here only when they explicitly name this exact model as their source. Check each repository for available files.

Main repository files ↗safetensors
No other explicitly linked versions recorded.

Link to this listing

Made or quantized this model? Paste one of these into your model card or README so people can find this listing.

Markdown

[Listed on Custom Clankers](https://customclankers.com/models/sao10k-llama-3-3-70b-vulpecula-r1/)

HTML

<a href="https://customclankers.com/models/sao10k-llama-3-3-70b-vulpecula-r1/">Listed on Custom Clankers</a>

Plain text links only: no image, no script, nothing loaded from this site.

Customize or build on this model

Listed from public GitHub data, not tested by us. Read the code and license before you run anything.

Tools to customize it further

Remove more refusals

Heretic ↗ ★ 36,457
Removes refusal behavior from transformer models automatically. It pairs directional ablation (abliteration) with an Optuna parameter search that tries to cut refusals while keeping the model close to the original, so you don't have to pick layers by hand.
OBLITERATUS ↗ ★ 8,949
A larger refusal-removal toolkit with several abliteration methods, usable locally or through a hosted Hugging Face Space. Check the telemetry setting before you run it: the README describes telemetry as opt-in, but on by default in the Spaces version.

Fine-tune it on your own data

Unsloth ↗ ★ 78,549
A desktop app and Python library for running and training models locally. Its README lists LoRA, QLoRA, full fine-tuning, DPO and GRPO, plus export to formats including GGUF, on Windows, Linux, WSL and macOS.
LlamaFactory ↗ ★ 75,720
Fine-tunes more than 100 language and vision-language models from a command line or a browser-based Web UI, aimed at training without writing code.

Make a smaller download (quantize)

llama.cpp ↗ ★ 132,313
Runs models locally and includes the tools behind most GGUF files. A conversion script turns Hugging Face weights into GGUF, and llama-quantize shrinks a high-precision GGUF to lower-bit formats, which can cost some accuracy.
MLX LM ↗ ★ 7,470
Apple's MLX package for running and fine-tuning models on Apple silicon. It can quantize models and upload them to the Hugging Face Hub, and supports LoRA and full fine-tuning.

Merge it with another model

mergekit ↗ ★ 7,415
Merges pretrained models with many merge methods. It works out-of-core, so merges can run on a CPU or with as little as 8 GB of VRAM, and it also handles LoRA extraction and mixture-of-experts merges.

Learn the techniques

LLM Course ↗ ★ 84,684
A free course with roadmaps and Colab notebooks. Its notebooks include one-click tools for merging with MergeKit, quantizing to GGUF, GPTQ, EXL2, AWQ and HQQ, and automatic abliteration with custom datasets.

Compare all tools →

Source provenance

Downloads: 106
Likes: 66
Metadata checked: 2026-10-25 · 16:11:04 UTC
First seen in index: 2026-10-25 · 16:11:04 UTC
Source revision: 12d7254ab9a5ce21905f59f341a3d2a2b3e62fd5

Source-card excerpt

Source model card excerpt: 🦊 L3.3-70B-Vulpecula 🌸 Hi hi! 🌟 This is a collaboration work between GradientPutri and Sao10K . This is a passion project of mine spanning

From Sao10K/Llama-3.3-70B-Vulpecula-r1. Creator model-card excerpt, not an independent review.

Inspect before you run

Low refusal is not proof of correctness. Outputs can be inaccurate, harmful, or biased. Review the model card, license and files yourself. A license tag is not commercial-use advice. Parameter counts are source metadata; no hardware compatibility is implied.

This directory does not host or redistribute model weights. Refusal profiles are measured by us on a fixed prompt set; all other behavior claims are source-reported.

Collection rules & limitations ↗