Upload folder using huggingface_hub

README.RU.md

@@ -13,7 +13,6 @@ tags:
 - quantized
 - auto-round
 - w4a16
-- uncensored
 - heretic
 - mtp
 - 256k-context

README.md

@@ -13,19 +13,10 @@ tags:
 - quantized
 - auto-round
 - w4a16
-- uncensored
 - heretic
 - mtp
 - 256k-context
 pipeline_tag: image-text-to-text
-datasets:
-- armand0e/claude-fable-5-claude-code
-- greghavens/fable-5-coding-and-debugging-traces
-- >-
-  CodeFlame/Qwen3.8-GLM5.2-Kimi-K3-GPT5.6-Gemini-3.1-Claude-Fable5-Mythos5-distillation
-- Crownelius/Complete-FABLE.5-traces-2M
-base_model:
-- Qwen/Qwen3.8-27B
 ---
 
 # goldhub/Qwen3.8-27B-INT4-W4A16-AutoRound
@@ -41,7 +32,7 @@ Aligned for unrestricted, "Heretic" level reasoning, this model bypasses standar
 - **W4A16 AutoRound Quantization**: 4-bit weights, 16-bit activations. Group size 32, symmetric quantization, 1000 iterations for optimal calibration.
 - **Smart Layer Preservation**: Critical layers (Vision encoders, `linear_attn` projections, and embeddings) are explicitly kept in FP16/BF16 to prevent multimodal degradation and attention collapse.
 - **MTP Ready**: Full support for Multi-Token Prediction for blazing-fast inference speeds.
-- **Uncensored / Heretic Alignment**: Excels in deep reasoning, creative writing, and unfiltered philosophical exploration without preachy refusals.
+- **Semi-Uncensored / Heretic Alignment**: Excels in deep reasoning, creative writing, and unfiltered philosophical exploration without preachy refusals.
 
 ## 📊 Benchmark & Evaluation Highlights
 
@@ -204,4 +195,5 @@ python qwen_quantization.py \
 
 ---
 
-**Made with 🔥 by Goldhub. Use responsibly. Think deeply.**
+**Made with 🔥 by Goldhub. Use responsibly. Think deeply.**
+