Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,13 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+assets/comparison_composition_1.png filter=lfs diff=lfs merge=lfs -text
+assets/comparison_composition_2.png filter=lfs diff=lfs merge=lfs -text
+assets/comparison_editing_1.png filter=lfs diff=lfs merge=lfs -text
+assets/comparison_editing_2.png filter=lfs diff=lfs merge=lfs -text
+assets/comparison_t2i_1.png filter=lfs diff=lfs merge=lfs -text
+assets/comparison_t2i_2.png filter=lfs diff=lfs merge=lfs -text
+assets/reference_composition_1.png filter=lfs diff=lfs merge=lfs -text
+assets/reference_composition_2.png filter=lfs diff=lfs merge=lfs -text
+assets/reference_editing_1.png filter=lfs diff=lfs merge=lfs -text
+assets/reference_editing_2.png filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,168 @@
+---
+library_name: diffusers
+license: apache-2.0
+base_model: black-forest-labs/FLUX.2-klein-4B
+base_model_relation: quantized
+tags:
+  - flux
+  - flux2
+  - quantized
+  - int8
+  - transformer
+  - nvidia-modelopt
+pipeline_tag: text-to-image
+---
+
+# FLUX.2-klein-4b-INT8-transformer-quants
+
+INT8 (W8A8) quantization variants for [FLUX.2-klein-4B](https://huggingface.co/black-forest-labs/FLUX.2-klein-4B) (step-distilled 4B parameters).
+
+This repository contains multiple INT8 quantization variants for experimentation and comparison.
+
+> **Status**: Static/Max variants (`int8-per-row`, `int8-per-tensor`) are available now.
+> SmoothQuant variants are pending and will be added when ready.
+
+| Variant | Algorithm | Scale Mode | Status | Checkpoint |
+|---------|-----------|------------|--------|------------|
+| int8-per-row | static | per-row | ✅ Available | `flux-2-klein-4b-int8-per-row.safetensors` |
+| int8-per-tensor | static | per-tensor | ✅ Available | `flux-2-klein-4b-int8-per-tensor.safetensors` |
+| int8-smoothquant-per-row | smoothquant | per-row | 🔜 Pending | `flux-2-klein-4b-int8-smoothquant-per-row.safetensors` |
+| int8-smoothquant-per-tensor | smoothquant | per-tensor | 🔜 Pending | `flux-2-klein-4b-int8-smoothquant-per-tensor.safetensors` |
+
+## Quantization Details
+
+All variants use [NVIDIA TensorRT Model Optimizer (ModelOpt)](https://github.com/NVIDIA/TensorRT-Model-Optimizer)
+INT8 (W8A8) quantization:
+
+| Property | Value |
+|----------|-------|
+| Framework | NVIDIA ModelOpt |
+| Calibration | 768 prompts (256 T2I, 256 editing, 256 composition), 4 steps each |
+| Weight Quantization | INT8 symmetric — per-row or per-tensor depending on variant |
+| Activation Quantization | Dynamic per-row (quantized on-the-fly at inference, one scale per token) |
+| Preserved Layers | Embedder layers (time_embed, context_embedder, x_embedder) and output projection kept in BF16 |
+
+### Algorithm × Scale Mode
+
+| | **Per-Row** | **Per-Tensor** |
+|---|---|---|
+| **Static (Max)** | `int8-per-row` ✅ | `int8-per-tensor` ✅ |
+| **SmoothQuant** | `int8-smoothquant-per-row` 🔜 | `int8-smoothquant-per-tensor` 🔜 |
+
+**Algorithm:**
+- **Static (Max)**: Standard INT8 quantization with calibrated min/max ranges
+- **SmoothQuant** *(pending)*: Migrates quantization difficulty from activations to weights for better accuracy
+
+**Scale Mode:**
+- **Per-Row**: Independent scale per output channel (finer granularity, higher accuracy)
+- **Per-Tensor**: Single scale per tensor (faster, lower memory, slightly reduced accuracy)
+
+> **Note**: In all variants, input activations are always quantized **dynamically per-row** at inference time
+> (one scale per token). The scale mode above refers to the **weight** quantization granularity.
+
+## Evaluation Results
+
+Compared against BF16 baseline using identical prompts, seeds, and resolution.
+
+### Overall Metrics
+
+| Variant | CLIP ↑ | LPIPS ↓ | PSNR ↑ | MSE ↓ | FID ↓ |
+|---------|--------|---------|--------|-------|-------|
+| BF16 (baseline) | 0.6518 | — | — | — | — |
+| FP8 (reference) | 0.6522 | 0.0356 | 27.60 | 210.14 | 22.94 |
+| int8-per-row | 0.6526 | 0.0253 | 29.23 | 142.27 | 15.76 |
+| int8-per-tensor | 0.6517 | 0.0447 | 26.25 | 254.55 | 24.83 |
+| int8-smoothquant-per-row | — | — | — | — | — |
+| int8-smoothquant-per-tensor | — | — | — | — | — |
+
+### Text-to-Image
+
+| Variant | CLIP ↑ | LPIPS ↓ | PSNR ↑ |
+|---------|--------|---------|--------|
+| FP8 (reference) | 0.6452 | 0.0480 | 24.43 |
+| int8-per-row | 0.6458 | 0.0343 | 25.89 |
+| int8-per-tensor | 0.6466 | 0.0621 | 23.23 |
+| int8-smoothquant-per-row | — | — | — |
+| int8-smoothquant-per-tensor | — | — | — |
+
+> Dramatic chiaroscuro portrait of a cellist mid-performance, single spotlight from above, instrument bow caught in motion blur, concert hall darkness
+
+![Text-to-Image 1 — BF16 vs FP8 vs INT8](assets/comparison_t2i_1.png)
+
+> Stained glass window design depicting the four elements, lead came outlines, rich jewel tones of ruby, sapphire, emerald, and topaz
+
+![Text-to-Image 2 — BF16 vs FP8 vs INT8](assets/comparison_t2i_2.png)
+
+### Editing
+
+| Variant | CLIP ↑ | LPIPS ↓ | PSNR ↑ |
+|---------|--------|---------|--------|
+| FP8 (reference) | 0.6467 | 0.0178 | 32.28 |
+| int8-per-row | 0.6465 | 0.0150 | 33.69 |
+| int8-per-tensor | 0.6447 | 0.0256 | 30.21 |
+| int8-smoothquant-per-row | — | — | — |
+| int8-smoothquant-per-tensor | — | — | — |
+
+> **Base:** A bicycle leaning against a stone wall in a village
+>
+> **Edit:** Transform the village into an underwater coral reef scene, the bicycle covered in barnacles and sea anemones, fish swimming around
+
+![Editing 1 — reference](assets/reference_editing_1.png)
+
+![Editing 1 — BF16 vs FP8 vs INT8](assets/comparison_editing_1.png)
+
+> **Base:** A food truck parked on a city street at noon
+>
+> **Edit:** Change the street to a Venice canal with the food truck floating on a gondola platform, evening golden hour lighting
+
+![Editing 2 — reference](assets/reference_editing_2.png)
+
+![Editing 2 — BF16 vs FP8 vs INT8](assets/comparison_editing_2.png)
+
+### Composition
+
+| Variant | CLIP ↑ | LPIPS ↓ | PSNR ↑ |
+|---------|--------|---------|--------|
+| FP8 (reference) | 0.6649 | 0.0410 | 26.08 |
+| int8-per-row | 0.6654 | 0.0267 | 28.10 |
+| int8-per-tensor | 0.6638 | 0.0465 | 25.32 |
+| int8-smoothquant-per-row | — | — | — |
+| int8-smoothquant-per-tensor | — | — | — |
+
+> Create a zen garden where the raked sand patterns flow into and around a giant ramen bowl as the central stone
+
+![Composition 1 — reference](assets/reference_composition_1.png)
+
+![Composition 1 — BF16 vs FP8 vs INT8](assets/comparison_composition_1.png)
+
+> A clockwork mechanical wolf made of brass gears howling at the full moon on the snowy ridge, steam rising from its joints
+
+![Composition 2 — reference](assets/reference_composition_2.png)
+
+![Composition 2 — BF16 vs FP8 vs INT8](assets/comparison_composition_2.png)
+
+## Usage
+
+> **🚧 Code release coming soon.** A pip-installable loader library is in preparation.
+
+In the meantime, these checkpoints can be tested with ComfyUI using the
+[ComfyUI-Flux2-INT8](https://github.com/BobJohnson24/ComfyUI-Flux2-INT8) custom node.
+Per-row quantization support is available via
+[PR #24](https://github.com/BobJohnson24/ComfyUI-Flux2-INT8/pull/24).
+
+## Technical Details
+
+| Property | Value |
+|----------|-------|
+| Base Model | [FLUX.2-klein-4B](https://huggingface.co/black-forest-labs/FLUX.2-klein-4B) |
+| Parameters | 4B |
+| Quantization | INT8 (W8A8) via NVIDIA ModelOpt |
+| Calibration | 768 prompts (256 per task), 4 steps each |
+| Activation Quantization | Dynamic per-row (quantized on-the-fly at inference) |
+| Preserved Layers | Embedder layers and output projection kept in BF16 |
+| Inference Steps | 4 |
+| Guidance Scale | 1.0 |
+
+## License
+
+This model inherits the license from the base model: **[Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0)**.

flux-2-klein-4b-int8-per-row.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:992b62763f59fb41626ca6a8de3ede910c0216d155d447a2dcb0f5334325f936
+size 4074429688

flux-2-klein-4b-int8-per-tensor.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fbd6cbc80f66af842d7cfe64eb10d55a1322c12bee5bf2d2e089666c5784d68a
+size 4070618904