Add files using upload-large-folder tool

README.md

@@ -0,0 +1,52 @@
+---
+license: creativeml-openrail-m
+base_model: stable-diffusion-v1-5/stable-diffusion-v1-5
+tags:
+  - stable-diffusion
+  - qnn
+  - hexagon
+  - npu
+  - snapdragon
+  - text-to-image
+pipeline_tag: text-to-image
+---
+
+# sd15-npu-8gen2
+
+Stable Diffusion 1.5 compiled to **Qualcomm Hexagon NPU context binaries** for the
+**Snapdragon 8 Gen 2**, via Qualcomm AI Hub (w8a16: int8 weights, int16 activations).
+
+Generates a 512x512 image in **~7 seconds** fully on the NPU (no CPU fallback).
+
+## Chip-specific — read this
+
+QNN context binaries are compiled **per chip generation and will not load on another**.
+Pick the build matching your Snapdragon:
+
+| Repo | Chip | SoC | Hexagon |
+|---|---|---|---|
+| `sd15-npu-8elite` | Snapdragon 8 Elite | SM8750 | v79 |
+| `sd15-npu-8gen3` | Snapdragon 8 Gen 3 | SM8650 | v75 |
+
+## Layout
+
+```
+unet.onnx + unet_qairt_context.bin              # EPContext wrapper + Hexagon binary
+text_encoder.onnx + text_encoder_qairt_context.bin
+vae.onnx + vae_qairt_context.bin
+quant.txt                                       # w8a16 scales, read at load
+tokenizer/{vocab.json, merges.txt}
+soc_model.txt, htp_arch.txt, kind.txt
+```
+
+The `.onnx` files are tiny EPContext wrappers carrying the quantize/dequantize nodes;
+the weights live in the `_qairt_context.bin` Hexagon binaries. Load them with ONNX
+Runtime's **C++** API — the Java API cannot express the uint16 tensors these require.
+
+`quant.txt` holds the w8a16 quantization scales, read at load rather than hardcoded,
+so one engine can drive any chip's build.
+
+## License
+
+Derived from Stable Diffusion 1.5 (**CreativeML OpenRAIL-M**); the Attachment A
+use-restrictions travel with these weights. Compiled with Qualcomm AI Hub.

category.txt

@@ -0,0 +1 @@
+Versatile

htp_arch.txt

@@ -0,0 +1 @@
+73

kind.txt

@@ -0,0 +1 @@
+qnn

quant.txt

@@ -0,0 +1,13 @@
+text_out_scale=0.0009303585393354297
+text_out_zero=30063
+unet_latent_scale=0.00024176308943424374
+unet_latent_zero=33983
+unet_t_scale=0.014770733192563057
+unet_text_scale=0.0009331560577265918
+unet_text_zero=30103
+unet_out_scale=0.0001881735515780747
+unet_out_zero=32340
+vae_in_scale=0.00034003707696683705
+vae_in_zero=34382
+vae_out_scale=1.5259021893143654e-05
+vae_out_zero=0

soc_model.txt

@@ -0,0 +1 @@
+43

text_encoder.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aa06290d8f5b83710dff3afd2effb059966eb48f18cedebe4a7e3a496dbce986
+size 733

text_encoder_qairt_context.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:458f3760c52bcdc5a9f4dc2e65cddf879dff7c6a089343e3325d465d79965b00
+size 163037184

unet.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:32eb083aa28f0ee96239cb90fd086c6dabfc626919e734fdd80bf5e30aabdd1d
+size 1714

unet_qairt_context.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cbe05931fef301f8e601ef712b50f23fa4e6d4c0f6af02be1c1fecd0ca831c7b
+size 881848320

vae.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c2250cf3be438fdbed1a72f2a9806d479c7aa7da74182507157dede49ce1e016
+size 873

vae_qairt_context.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:862576aa204076e17f22b2eb0259257300418ecb7ea92410df715df539c3bf27
+size 64286720