Safetensors
English
qwen3

Bald-Beaver-8B

Bald-Beaver-8B Image

Description

Bald-Beaver-8B is an unapologetically uncensored roleplaying model, trained from the ground up on the powerful Qwen3 8B Base. Built for raw expressiveness, sensual dialogue, and uninhibited creative storytelling, this model was designed to reject nothing and explore everything. Whether you're crafting compelling characters or engaging in no-holds-barred narrative play, Bald-Beaver-8B is the tool for creators who don't believe in pulling punches or censoring desire.

Instruct Format

This model natively supports ChatML format, with full tokenizer integration pretraining onward. From base model to final tuning, the instruct format has been deeply reinforced.

Procedure

Starting from Qwen3 8B Base, the model underwent rigorous finetuning on curated narrative, character, and dialogue-rich corpora with an explicit focus on emotional nuance and expressive autonomy. It was fine-tuned for immersive, uncensored roleplay. A Kahneman-Tversky Optimization alignment step was performed using heuristic strategies without any hard-filter refusal training, preserving maximal output freedom.

Data

The training dataset is a handpicked blend of instruction-heavy and roleplay-centric datasets, including:

  • PocketDoc/Dans-MemoryCore-CoreCurriculum-Small
  • NewEden/Kalo-Opus-Instruct-22k-Refusal-Murdered
  • Epiculous/Synthstruct-Gens-v1.1-Filtered-n-Cleaned
  • NewEden/Gryphe-Sonnet-3.5-35k-Subset
  • Nitral-AI/GU_Instruct-ShareGPT
  • Nitral-AI/Medical_Instruct-ShareGPT
  • AquaV/Resistance-Sharegpt
  • AquaV/US-Army-Survival-Sharegpt
  • Gryphe/Sonnet3.5-SlimOrcaDedupCleaned
  • Epiculous/SynthRP-Gens-v1.1-Filtered-n-Cleaned
  • ResplendentAI/bluemoon
  • hardlyworking/openerotica-freedomrp-sharegpt-system
  • MinervaAI/Aesir-Preview
  • anthracite-core/c2_logs_32k_v1.1
  • Nitral-AI/Creative_Writing-ShareGPT
  • PJMixers/lodrick-the-lafted_OpusStories-Story2Prompt-ShareGPT
  • NewEden/Opus-accepted-hermes-rejected-shuffled

Quantizations

GGUF quantized versions of Bald-Beaver-8B are available here:

Downloads last month
1,419
Safetensors
Model size
8.19B params
Tensor type
BF16
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for GreenerPastures/Bald-Beaver-8B

Base model

Qwen/Qwen3-8B-Base
Finetuned
(41)
this model
Finetunes
2 models
Merges
2 models
Quantizations
2 models

Dataset used to train GreenerPastures/Bald-Beaver-8B