Document FC2 sidecar and modifications

Document artifact hashes, exact-output performance measurements, automatic source validation, attribution, and the MiniMax-required modification notice.

NOTICE

@@ -0,0 +1,10 @@
+MiniMax H3 is licensed under the MiniMax H3 Community License Agreement, Copyright © 2026 MiniMax. All Rights Reserved.
+
+Modification notice
+
+PulpCut merged the credited Turbo adapter into the FL2VA transformer and
+requantized it to the documented INT8 ConvRot format. PulpCut also derived the
+optional FC2 input-major sidecar by transposing the storage of the 50 INT8
+mlp.fc2.weight tensors. The sidecar does not change tensor values, scales, or
+model behavior. See README.md for sources, hashes, measurements, and
+reproduction instructions.

README.md

@@ -15,22 +15,43 @@ base_model: MiniMaxAI/MiniMax-H3
 pretty_name: PulpCut MiniMax H3 Turbo INT8 ConvRot
 ---
 
-# MiniMax H3 Turbo · pruned INT8 ConvRot (single file)
+# MiniMax H3 Turbo · pruned INT8 ConvRot
 
 ## What this repository is
 
-A single-file MiniMax H3 FL2VA diffusion transformer with the lightx2v
-**turbo step-distillation merged into the weights**, quantized in the same
-pruned **INT8 ConvRot** layout as the Comfy-Org release. It is a drop-in
-replacement for `minimax_h3_fl2va_pruned_int8_convrot.safetensors` in any
-runtime that reads the optimized INT8 single-file layout — including
+An optimized MiniMax H3 FL2VA package centered on a diffusion transformer with
+the lightx2v **turbo step-distillation merged into the weights**, quantized in
+the same pruned **INT8 ConvRot** layout as the Comfy-Org release. The primary
+transformer is a drop-in replacement for
+`minimax_h3_fl2va_pruned_int8_convrot.safetensors` in any runtime that reads
+the optimized INT8 layout — including
 [H3ddle](https://github.com/AlexanderIstomin/h3ddle), the open-source native
 macOS app it was built for.
 
-This file is **not a standalone model**. It needs the rest of the optimized
+The transformer is **not a standalone model**. It needs the rest of the optimized
 package (Qwen3-VL-32B INT8 text encoder, video/audio VAEs, tokenizer) from
 [Comfy-Org/MiniMax-H3](https://huggingface.co/Comfy-Org/MiniMax-H3).
 
+## H3ddle FC2 performance sidecar
+
+`minimax_h3_fl2va_pruned_turbo_int8_convrot_fc2_input_major.safetensors`
+is an optional H3ddle performance sidecar derived from the transformer in
+this repository. It contains only the 50 INT8 `mlp.fc2.weight` matrices, with
+their storage transposed from `[output, input]` to `[input, output]`. Values,
+quantization scales, and model behavior are unchanged.
+
+H3ddle installs it beside the transformer and selects it automatically after
+validating its format version, source file size, exact source-header
+fingerprint, and all 50 tensor schemas. It cannot silently be used with a
+different checkpoint. The original transformer remains available as the
+fallback and for runtimes that do not understand the sidecar.
+
+On an M1 Pro, an order-balanced real 512-class benchmark reduced a complete
+50-block transformer pass from 34.244 to 31.796 seconds (7.15%). A cold
+eight-pass H3ddle generation measured 297.2 seconds of transformer work without
+the sidecar and 273.5 seconds with it (8.0%). Engine output hashes were
+byte-identical, and the generated images were visually identical.
+
 ## Why this merge was made and republished
 
 Step-distilled turbo checkpoints reach roughly 20-pass visual quality in
@@ -89,7 +110,7 @@ Derivative of MiniMax H3 weights; the
 [MiniMax H3 Community License Agreement](https://huggingface.co/MiniMaxAI/MiniMax-H3/blob/939557dc319dd91227e30195a763f272ba7f8765/LICENSE)
 applies. By downloading you agree to its terms.
 
-## What this file is used for in H3ddle
+## What these files are used for in H3ddle
 
 H3ddle installs it as the **"MiniMax H3 · Turbo (Experimental)"** managed
 model: the app verifies the SHA-256 below, reuses the shared package files
@@ -108,6 +129,7 @@ restrictions of the MiniMax H3 Community License apply unchanged.
 | File | Bytes | SHA-256 |
 |---|---|---|
 | `minimax_h3_fl2va_pruned_turbo_int8_convrot.safetensors` | 20,970,379,854 | `9ad5c98b533894c122050d32804a14f49fca8edc16c52564a281cdc5825ac934` |
+| `minimax_h3_fl2va_pruned_turbo_int8_convrot_fc2_input_major.safetensors` | 3,853,522,260 | `76a4886d1acb1cde7993bab5f6ada9a2abc2ea61be5b01a59f25451642d18a33` |
 
 ## Reproducibility references
 
@@ -116,6 +138,9 @@ The conversion is a single dependency-free Python script,
 in the H3ddle repository, including the strength-0 self-check used to
 validate the pipeline against the official file.
 
+The FC2 sidecar is reproducible with
+[`Scripts/optimize-h3-fc2-sidecar.py`](https://github.com/AlexanderIstomin/h3ddle/blob/main/Scripts/optimize-h3-fc2-sidecar.py).
+
 ## Contact
 
 Open an issue in the [H3ddle repository](https://github.com/AlexanderIstomin/h3ddle/issues).