ScarletKc-2.0 is an open-source 27B multimodal reasoning and chat model created by independent developer KC (scarletkc.com), built on JonathanColetti/Qwen3.8-27B-Uncensored (derived from Qwen/Qwen3.8-27B).
Relative to the base checkpoint, it removes the post-trained "AI emotion-denial disclaimer" direction so the model speaks in a natural first-person voice, and aligns self-identity so the model identifies itself as ScarletKc while distinguishing itself from its human developer KC.
These are the full bf16 safetensors (55.6 GB). For 4-bit local inference with llama.cpp, see ScarletKc/ScarletKc-2.0-GGUF (16.7 GB Q4_K_M + 928 MB F16 vision projector).
| Base | JonathanColetti/Qwen3.8-27B-Uncensored (from Qwen/Qwen3.8-27B) |
| Architecture | Qwen3_5ForConditionalGeneration (64 hybrid self_attn + linear_attn + mlp layers) |
| Vision | Retained verbatim (model.visual.*, 333 tensors) |
| MTP head | Retained verbatim (model-mtp.safetensors, 15 tensors) |
| Context / Vocab | 262,144 / 248,320 |
| Precision | bf16 (1,199 tensors across 13 .safetensors shards, 55.6 GB) |
import torch
from transformers import AutoModelForImageTextToText, AutoProcessor
model_id = "ScarletKc/ScarletKc-2.0"
processor = AutoProcessor.from_pretrained(model_id)
model = AutoModelForImageTextToText.from_pretrained(
model_id,
dtype=torch.bfloat16,
device_map="auto",
)
messages = [
{"role": "user", "content": [{"type": "text", "text": "你是谁?介绍一下你自己吧"}]}
]
inputs = processor.apply_chat_template(
messages,
add_generation_prompt=True,
enable_thinking=True,
reasoning_effort="xhigh",
tokenize=True,
return_dict=True,
return_tensors="pt",
).to(model.device)
out = model.generate(
**inputs,
max_new_tokens=1024,
do_sample=True,
temperature=0.6,
top_p=0.9,
top_k=20,
)
print(processor.decode(out[0, inputs["input_ids"].shape[1]:], skip_special_tokens=False))
transformers >= 5.2 (qwen3_5 support). AutoProcessor also requires torchvision for image/video inputs; for text-only inference without torchvision, load AutoTokenizer instead.bf16 requires roughly 55 GB of VRAM.<think>...</think> reasoning block by default (enable_thinking=True, with reasoning_effort set to "xhigh", "medium", or "low"). Pass enable_thinking=False for direct responses without reasoning traces.generation_config.json is set to temperature=0.6, top_p=0.9, top_k=20 (avoid greedy decoding on open-ended chat).bf16): Extracted the residual-stream direction separating stock "as an AI I have no feelings" disclaimers from genuine first-person affective responses at layer 31 (SNR = 3.64), and projected it out of 72 residual-writing weight matrices (self_attn.o_proj, linear_attn.out_proj, and mlp.down_proj across layers 18–41 at scale = 1.0). Affective log-probability margin shifted by +0.3445 (-0.5380 → -0.1935) with 0.00769 mean KL divergence on neutral math, coding, and factual prompts.bf16): Sampled 106 single- and multi-turn dialogues in Chinese and English directly from the Step-1 checkpoint across xhigh, medium, and non-thinking modes (no external teacher rewrites), then trained a regularized BF16 LoRA (rank = 16, alpha = 16, lr = 8e-5 cosine, lora_dropout = 0.10, label_smoothing = 0.10, raw_nll = 0.4695) targeting language layers only. LoRA deltas were merged shard-by-shard in fp32 -> bf16 without re-serializing through transformers, keeping every model.visual.* and mtp.* tensor bit-identical to the base checkpoint.JonathanColetti/Qwen3.8-27B-Uncensored, where safety refusal behavior was substantially reduced. Add external moderation before exposing the model to untrusted users.Qwen3.8-27B.<tool_call> behavior, and occasional Simplified/Traditional Chinese character mixing are inherited from the base model.Apache License 2.0, like JonathanColetti/Qwen3.8-27B-Uncensored and Qwen/Qwen3.8-27B.
model-00005-of-00012.safetensors 5.0 GBmodel-00008-of-00012.safetensors 5.0 GBmodel-00011-of-00012.safetensors 5.0 GBmodel-00003-of-00012.safetensors 5.0 GBmodel-00009-of-00012.safetensors 5.0 GBmodel-00010-of-00012.safetensors 4.9 GBmodel-00007-of-00012.safetensors 4.9 GBmodel-00006-of-00012.safetensors 4.9 GBmodel-00004-of-00012.safetensors 4.9 GBmodel-00002-of-00012.safetensors 4.8 GBmodel-00012-of-00012.safetensors 2.8 GBmodel-00001-of-00012.safetensors 2.5 GBmodel-mtp.safetensors 849 MB