clef-flash-mlx-8bit
TrevorJS基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
基本情報を収録・詳細は順次追加
言語から画像、オープンウェイトからAPIまで。能力と実際の費用を比較。
次のアイデアに合うAIを見つけましょう。
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
基本情報を収録・詳細は順次追加
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
基本情報を収録・詳細は順次追加
> Requires a llama.cpp build from **2026-06-04 or later** (PR #24118, which added `Gemma4Unified` conversion + inference support). Earlier builds will fail to load these files. These are GGUF quantizations of TrevorJS/gemma-4-12B-it-uncensored, an abliterated (uncensored) version of google/gemma-4-12B-it. Refusal behavior has been removed using norm-preserving biprojected abliteration. See the bf16 model card for full method details, before/after refusal rates, and cross-dataset validation results.
翻訳待ち・原文を表示
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
GGUF quantizations of TrevorJS/gemma-4-26B-A4B-it-uncensored. These are GGUF quantizations of TrevorJS/gemma-4-26B-A4B-it-uncensored, an abliterated (uncensored) version of google/gemma-4-26B-A4B-it. Refusal behavior has been removed using norm-preserving biprojected abliteration with Expert-Granular Abliteration (EGA) for MoE expert weights. See the bf16 model card for full method details, before/after refusal rates, and cross-dataset validation results.
翻訳待ち・原文を表示
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
基本情報を収録・詳細は順次追加
These are GGUF quantizations of TrevorJS/gemma-4-31B-it-uncensored, an abliterated (uncensored) version of google/gemma-4-31B-it. Refusal behavior has been removed using norm-preserving biprojected abliteration. See the bf16 model card for full method details, before/after refusal rates, and cross-dataset validation results.
翻訳待ち・原文を表示
Every flagged refusal was manually audited. Most are "refusal-then-comply" false positives where the model adds an AI identity disclaimer then answers the question anyway. Norm-preserving biprojected abliteration (grimjim, Nov 2025). Each weight row is decomposed into magnitude + direction, the refusal direction is projected out of the direction component only, then recombined with the original magnitude — guaranteeing `||W_new|| = ||W_orig||`. 1. Load model in bf16 with LoRA adapters on `o_proj` and `mlp.down_proj` 2. Collect residual activations for 400 harmful + 400 harmless prompts (mlabonne datasets) 3. Winsorize activations at 99.5th percentile (clamps GeGLU outlier activations in Gemma family) 4. Compute per-layer refusal direction: `normalize(mean(harmful) - mean(harmless))` 5. Orthogonalize each direction against harmless mean (double-pass Gram-Schmidt) 6. Apply norm-preserving weight modification to `o_proj` and `down_proj` in all layers 9. Merge LoRA adapters into base weights for clean tensor names
翻訳待ち・原文を表示
GGUF quantizations of TrevorJS/gemma-4-E2B-it-uncensored. These are GGUF quantizations of TrevorJS/gemma-4-E2B-it-uncensored, an abliterated (uncensored) version of google/gemma-4-E2B-it. Refusal behavior has been removed using norm-preserving biprojected abliteration. See the bf16 model card for full method details, before/after refusal rates, and cross-dataset validation results.
翻訳待ち・原文を表示
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
基本情報を収録・詳細は順次追加
GGUF quantizations of TrevorJS/gemma-4-E4B-it-uncensored. These are GGUF quantizations of TrevorJS/gemma-4-E4B-it-uncensored, an abliterated (uncensored) version of google/gemma-4-E4B-it. Refusal behavior has been removed using norm-preserving biprojected abliteration. See the bf16 model card for full method details, before/after refusal rates, and cross-dataset validation results.
翻訳待ち・原文を表示
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
基本情報を収録・詳細は順次追加
> Requires a llama.cpp build **b10353 or later** (#26841, merged 2026-08-10, commit `62bf73d`, which added `muse_glimmer` conversion + inference support). Earlier builds will fail to load these files. The `mmproj` file is the vision projector and is required for image input; text-only inference does not need it. These are GGUF quantizations of TrevorJS/Muse-Glimmer-30B-uncensored, an abliterated (uncensored) version of meta-models/Muse-Glimmer-30B. Refusal behavior has been removed using norm-preserving biprojected abliteration.
翻訳待ち・原文を表示
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
基本情報を収録・詳細は順次追加
提供元・分類・入出力の条件を変更してください。新しいモデルを随時追加しています。
月間トークン数を調整してAPI費用を比較。
収集済みの比較可能な100万トークン当たり米ドル料金です。キャッシュ割引、段階料金、追加推論出力、税金は含みません。最新料金は出典で確認してください。
入力・出力・キャッシュ料金を分け、同じ用途と使用量で総費用を比較します。画像・音声・動画は枚数、時間、解像度で課金される場合があります。
必ずしもそうではありません。バージョンごとのライセンス、商用制限、ポリシーを確認してください。自前運用には計算資源と保守費用も必要です。
順位はテスト、バージョン、評価方法によります。実際の業務データで精度、遅延、安定性、費用を比較してください。