license: apache-2.0
base_model: anaaaaaa13421312/gemma-4-31B-it-abliterated-Japanese-RP
base_model_relation: quantized
library_name: llm-compressor
tags:
- compressed-tensors
- gemma4
- w8a8
- japanese
- roleplay
- not-for-all-audiences
gemma-4-31B-it-abliterated-Japanese-RP-INT8-W8A8
anaaaaaa13421312/gemma-4-31B-it-abliterated-Japanese-RP を llmcompressor.model_free_ptq(scheme="W8A8")でデータフリー量子化したチェックポイントです。
- Format:
compressed-tensors - Scheme:
W8A8 - Target: INT8 W8A8 (INT8 per-channel static weights + dynamic per-token activations) for TPU v5e/v6e and RunPod A40 (48GB Ampere).
- Unquantized (BF16 preserved):
lm_head, vision/audio modules,embed_tokens, norms