SDXL-Lightning 4-step int8 ONNX (sharded UNet, 2 chunks)

.gitattributes

@@ -33,3 +33,8 @@ 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
+text_encoder/model.onnx_data filter=lfs diff=lfs merge=lfs -text
+text_encoder_2/model.onnx_data filter=lfs diff=lfs merge=lfs -text
+unet/model_int8.onnx_data filter=lfs diff=lfs merge=lfs -text
+unet/model_int8.onnx_data_1 filter=lfs diff=lfs merge=lfs -text
+vae_decoder/model.onnx_data filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,96 @@
+---
+license: openrail++
+tags:
+  - stable-diffusion-xl
+  - sdxl
+  - sdxl-lightning
+  - onnx
+  - onnxruntime-web
+  - text-to-image
+  - distilled
+  - fp16
+library_name: optimum
+pipeline_tag: text-to-image
+---
+
+# SDXL-Lightning 4-step (ONNX, fp16, ORT-Web compatible)
+
+ONNX export of [ByteDance/SDXL-Lightning](https://huggingface.co/ByteDance/SDXL-Lightning)
+4-step UNet merged into [stabilityai/stable-diffusion-xl-base-1.0](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0),
+with the VAE decoder replaced by [madebyollin/sdxl-vae-fp16-fix](https://huggingface.co/madebyollin/sdxl-vae-fp16-fix)
+for fp16 stability. Layout is a diffusers-style pipeline with per-subfolder
+ONNX files and bundled tokenizers, intended for in-browser inference via
+ONNX Runtime Web (WebGPU EP).
+
+## Contents
+
+| Subfolder | Size | Notes |
+|---|---|---|
+| `text_encoder/` | 236 MB | CLIP-L/14, fp16 |
+| `text_encoder_2/` | 1.3 GB | CLIP-G/14 (OpenCLIP-ViT-bigG-14), fp16 |
+| `unet/` | 4.8 GB | SDXL-Lightning 4-step UNet, fp16 |
+| `vae_decoder/` | 95 MB | sdxl-vae-fp16-fix decoder, fp16 |
+| `tokenizer/`, `tokenizer_2/` | small | Fast tokenizers (`tokenizer.json` present) |
+| `scheduler/` | small | EulerDiscrete, `timestep_spacing='trailing'` |
+| `model_index.json` | small | Diffusers pipeline manifest |
+
+Total: ~6.5 GB.
+
+## Recommended usage
+
+Designed for 4 denoising steps, classifier-free guidance disabled (CFG = 1.0).
+Guidance > 1 breaks Lightning. Recommended scheduler is the bundled
+`EulerDiscreteScheduler` with `timestep_spacing="trailing"`.
+
+The export targets ORT-Web's WebGPU EP. The CPU EP passes a sanity check
+locally; the WebGPU EP is the production target and tolerates a stricter
+op set. If a runtime op error appears in the browser, downgrade the opset
+or rebuild the export against a different toolchain version.
+
+## Production notes
+
+- `Resize` ops are kept at fp32 with auto-inserted casts at the boundary —
+  onnxconverter-common's default block list catches most cases, but the
+  `scales` Constant input was hand-patched back to fp32 after the converter
+  failed to insert casts for Constant-produced inputs (2 in UNet, 3 in VAE).
+- I/O dtypes are fp32 throughout (`keep_io_types=True`) so JavaScript callers
+  can feed unconverted fp32 tensors and read fp32 outputs.
+- The `vae_encoder/` from the original optimum export was dropped — Lightning
+  is text-to-image only.
+
+## Licenses
+
+This is a derivative work combining three upstream sources, each with its own
+license. All three are permissive but you should read them before commercial use.
+
+- **SDXL-Lightning UNet** — [ByteDance/SDXL-Lightning](https://huggingface.co/ByteDance/SDXL-Lightning)
+  is licensed under [CreativeML Open RAIL++-M](https://huggingface.co/ByteDance/SDXL-Lightning/blob/main/LICENSE.md).
+- **SDXL base-1.0** (everything except the UNet weights) — [stabilityai/stable-diffusion-xl-base-1.0](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0)
+  is licensed under [CreativeML Open RAIL++-M](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0/blob/main/LICENSE.md).
+- **VAE decoder** — [madebyollin/sdxl-vae-fp16-fix](https://huggingface.co/madebyollin/sdxl-vae-fp16-fix)
+  is MIT-licensed.
+
+The combined work is released under CreativeML Open RAIL++-M (the more
+restrictive of the upstream licenses).
+
+## How it was built
+
+Reproduction recipe (CPU-only Windows box):
+
+1. Construct the SDXL UNet from `stabilityai/stable-diffusion-xl-base-1.0`
+   config and load `sdxl_lightning_4step_unet.safetensors` from
+   `ByteDance/SDXL-Lightning` into it.
+2. Save the merged pipeline as a full diffusers pipeline.
+3. `optimum-cli export onnx --task stable-diffusion-xl --framework pt` →
+   per-subfolder ONNX at fp32 (~13 GB).
+4. Convert UNet + both text encoders to fp16 in place via
+   `onnxconverter-common.float16.convert_float_to_float16` with a custom
+   post-pass that reverts Resize-feeding Constants back to fp32.
+5. Replace the original VAE decoder with a fresh ONNX export of
+   `madebyollin/sdxl-vae-fp16-fix`, fp16-converted with the same post-pass.
+6. Build fast tokenizers (`tokenizer.json`) from the slow-tokenizer files
+   optimum-cli dropped, since transformers.js v3 has no slow-tokenizer fallback.
+
+Built with `torch==2.4.1+cpu`, `optimum[exporters]==1.23.3`,
+`transformers==4.45.2`, `diffusers==0.30.3`, `onnx==1.17.0`,
+`onnxruntime==1.20.1`, `onnxconverter-common==1.14.0`.

model_index.json

@@ -0,0 +1,42 @@
+{
+  "_class_name": "StableDiffusionXLPipeline",
+  "_diffusers_version": "0.30.3",
+  "_name_or_path": "C:/ml-export/sdxl-lightning-4step/sdxl-lightning-4step-merged",
+  "feature_extractor": [
+    null,
+    null
+  ],
+  "force_zeros_for_empty_prompt": true,
+  "image_encoder": [
+    null,
+    null
+  ],
+  "scheduler": [
+    "diffusers",
+    "EulerDiscreteScheduler"
+  ],
+  "text_encoder": [
+    "transformers",
+    "CLIPTextModel"
+  ],
+  "text_encoder_2": [
+    "transformers",
+    "CLIPTextModelWithProjection"
+  ],
+  "tokenizer": [
+    "transformers",
+    "CLIPTokenizer"
+  ],
+  "tokenizer_2": [
+    "transformers",
+    "CLIPTokenizer"
+  ],
+  "unet": [
+    "diffusers",
+    "UNet2DConditionModel"
+  ],
+  "vae": [
+    "diffusers",
+    "AutoencoderKL"
+  ]
+}

scheduler/scheduler_config.json

@@ -0,0 +1,23 @@
+{
+  "_class_name": "EulerDiscreteScheduler",
+  "_diffusers_version": "0.30.3",
+  "beta_end": 0.012,
+  "beta_schedule": "scaled_linear",
+  "beta_start": 0.00085,
+  "clip_sample": false,
+  "final_sigmas_type": "zero",
+  "interpolation_type": "linear",
+  "num_train_timesteps": 1000,
+  "prediction_type": "epsilon",
+  "rescale_betas_zero_snr": false,
+  "sample_max_value": 1.0,
+  "set_alpha_to_one": false,
+  "sigma_max": null,
+  "sigma_min": null,
+  "skip_prk_steps": true,
+  "steps_offset": 1,
+  "timestep_spacing": "trailing",
+  "timestep_type": "discrete",
+  "trained_betas": null,
+  "use_karras_sigmas": false
+}

text_encoder/config.json

@@ -0,0 +1,26 @@
+{
+  "_name_or_path": "C:/ml-export/sdxl-lightning-4step/sdxl-lightning-4step-merged\\text_encoder",
+  "architectures": [
+    "CLIPTextModel"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 0,
+  "dropout": 0.0,
+  "eos_token_id": 2,
+  "hidden_act": "quick_gelu",
+  "hidden_size": 768,
+  "initializer_factor": 1.0,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 77,
+  "model_type": "clip_text_model",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "output_hidden_states": true,
+  "pad_token_id": 1,
+  "projection_dim": 768,
+  "torch_dtype": "float32",
+  "transformers_version": "4.45.2",
+  "vocab_size": 49408
+}

text_encoder/model.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:21267b285b3e3ab33fc6d95610ce629a67e0bca01647c36e65e6cd04b6eaa6e7
+size 360437

text_encoder/model.onnx_data

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b61db0f4781c335e6bba50ab7471bbfe02aad41d050ff5935dcbb2d7724fda0
+size 246120960

text_encoder_2/config.json

@@ -0,0 +1,26 @@
+{
+  "_name_or_path": "C:/ml-export/sdxl-lightning-4step/sdxl-lightning-4step-merged\\text_encoder_2",
+  "architectures": [
+    "CLIPTextModelWithProjection"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 0,
+  "dropout": 0.0,
+  "eos_token_id": 2,
+  "hidden_act": "gelu",
+  "hidden_size": 1280,
+  "initializer_factor": 1.0,
+  "initializer_range": 0.02,
+  "intermediate_size": 5120,
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 77,
+  "model_type": "clip_text_model",
+  "num_attention_heads": 20,
+  "num_hidden_layers": 32,
+  "output_hidden_states": true,
+  "pad_token_id": 1,
+  "projection_dim": 1280,
+  "torch_dtype": "float32",
+  "transformers_version": "4.45.2",
+  "vocab_size": 49408
+}

text_encoder_2/model.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0cf6265108359805f27b16141209c83b9452d4ac66f7b626c6c6f046f648f484
+size 1047642

text_encoder_2/model.onnx_data

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:81330f8daecb663019d641658449d3d76205d90c2207356b9cccf1a427214c12
+size 1389319680

tokenizer/special_tokens_map.json

@@ -0,0 +1,30 @@
+{
+  "bos_token": {
+    "content": "<|startoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer/tokenizer_config.json

@@ -0,0 +1,30 @@
+{
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "49406": {
+      "content": "<|startoftext|>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "49407": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<|startoftext|>",
+  "clean_up_tokenization_spaces": true,
+  "do_lower_case": true,
+  "eos_token": "<|endoftext|>",
+  "errors": "replace",
+  "model_max_length": 77,
+  "pad_token": "<|endoftext|>",
+  "tokenizer_class": "CLIPTokenizer",
+  "unk_token": "<|endoftext|>"
+}

tokenizer_2/special_tokens_map.json

@@ -0,0 +1,30 @@
+{
+  "bos_token": {
+    "content": "<|startoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "!",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer_2/tokenizer_config.json

@@ -0,0 +1,38 @@
+{
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "!",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "49406": {
+      "content": "<|startoftext|>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "49407": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<|startoftext|>",
+  "clean_up_tokenization_spaces": true,
+  "do_lower_case": true,
+  "eos_token": "<|endoftext|>",
+  "errors": "replace",
+  "model_max_length": 77,
+  "pad_token": "!",
+  "tokenizer_class": "CLIPTokenizer",
+  "unk_token": "<|endoftext|>"
+}

unet/config.json

@@ -0,0 +1,73 @@
+{
+  "_class_name": "UNet2DConditionModel",
+  "_diffusers_version": "0.30.3",
+  "_name_or_path": "C:/ml-export/sdxl-lightning-4step/sdxl-lightning-4step-merged\\unet",
+  "act_fn": "silu",
+  "addition_embed_type": "text_time",
+  "addition_embed_type_num_heads": 64,
+  "addition_time_embed_dim": 256,
+  "attention_head_dim": [
+    5,
+    10,
+    20
+  ],
+  "attention_type": "default",
+  "block_out_channels": [
+    320,
+    640,
+    1280
+  ],
+  "center_input_sample": false,
+  "class_embed_type": null,
+  "class_embeddings_concat": false,
+  "conv_in_kernel": 3,
+  "conv_out_kernel": 3,
+  "cross_attention_dim": 2048,
+  "cross_attention_norm": null,
+  "down_block_types": [
+    "DownBlock2D",
+    "CrossAttnDownBlock2D",
+    "CrossAttnDownBlock2D"
+  ],
+  "downsample_padding": 1,
+  "dropout": 0.0,
+  "dual_cross_attention": false,
+  "encoder_hid_dim": null,
+  "encoder_hid_dim_type": null,
+  "flip_sin_to_cos": true,
+  "freq_shift": 0,
+  "in_channels": 4,
+  "layers_per_block": 2,
+  "mid_block_only_cross_attention": null,
+  "mid_block_scale_factor": 1,
+  "mid_block_type": "UNetMidBlock2DCrossAttn",
+  "norm_eps": 1e-05,
+  "norm_num_groups": 32,
+  "num_attention_heads": null,
+  "num_class_embeds": null,
+  "only_cross_attention": false,
+  "out_channels": 4,
+  "projection_class_embeddings_input_dim": 2816,
+  "resnet_out_scale_factor": 1.0,
+  "resnet_skip_time_act": false,
+  "resnet_time_scale_shift": "default",
+  "reverse_transformer_layers_per_block": null,
+  "sample_size": 128,
+  "time_cond_proj_dim": null,
+  "time_embedding_act_fn": null,
+  "time_embedding_dim": null,
+  "time_embedding_type": "positional",
+  "timestep_post_act": null,
+  "transformer_layers_per_block": [
+    1,
+    2,
+    10
+  ],
+  "up_block_types": [
+    "CrossAttnUpBlock2D",
+    "CrossAttnUpBlock2D",
+    "UpBlock2D"
+  ],
+  "upcast_attention": null,
+  "use_linear_projection": true
+}

unet/model_int8.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bbe5d37f958016dd180576264b5c667958665a27117c8013c5baff55f7bd195a
+size 6547812

unet/model_int8.onnx_data

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d78123a5c6c45524f2f5930018da87bbd1ed26df4238b7a588f346c2c231a205
+size 2000000000

unet/model_int8.onnx_data_1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da823220949e682698e6999aa2e16b10f8e40e0a564f92160cb7a1fd001b3e82
+size 1571911680

vae_decoder/config.json

@@ -0,0 +1,38 @@
+{
+  "_class_name": "AutoencoderKL",
+  "_diffusers_version": "0.30.3",
+  "_name_or_path": "C:/ml-export/sdxl-lightning-4step/sdxl-lightning-4step-merged\\vae",
+  "act_fn": "silu",
+  "block_out_channels": [
+    128,
+    256,
+    512,
+    512
+  ],
+  "down_block_types": [
+    "DownEncoderBlock2D",
+    "DownEncoderBlock2D",
+    "DownEncoderBlock2D",
+    "DownEncoderBlock2D"
+  ],
+  "force_upcast": true,
+  "in_channels": 3,
+  "latent_channels": 4,
+  "latents_mean": null,
+  "latents_std": null,
+  "layers_per_block": 2,
+  "mid_block_add_attention": true,
+  "norm_num_groups": 32,
+  "out_channels": 3,
+  "sample_size": 1024,
+  "scaling_factor": 0.13025,
+  "shift_factor": null,
+  "up_block_types": [
+    "UpDecoderBlock2D",
+    "UpDecoderBlock2D",
+    "UpDecoderBlock2D",
+    "UpDecoderBlock2D"
+  ],
+  "use_post_quant_conv": true,
+  "use_quant_conv": true
+}

vae_decoder/model.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d2dbbf47fba8ee8b1ec26c10b8a259d6e9870ec7d921df2b2d473fb97a2d0c03
+size 135522

vae_decoder/model.onnx_data

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fc3ff1700fd5f86a3e789aaadf647aa7742f2c739be6221ca692c5c92a21d39c
+size 98965248