Duplicate from Alissonerdx/BFS-Best-Face-Swap

Co-authored-by: Alisson <Alissonerdx@users.noreply.huggingface.co>

.gitattributes

@@ -0,0 +1,53 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz 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
+images/ComfyUI_04745_.png filter=lfs diff=lfs merge=lfs -text
+images/ComfyUI_04755_.png filter=lfs diff=lfs merge=lfs -text
+images/ComfyUI_04759_.png filter=lfs diff=lfs merge=lfs -text
+images/ComfyUI_04761_.png filter=lfs diff=lfs merge=lfs -text
+images/bfs_body_swap_v1_krea2_example_closeup.png filter=lfs diff=lfs merge=lfs -text
+images/bfs_body_swap_v1_krea2_example_kitchen.png filter=lfs diff=lfs merge=lfs -text
+images/bfs_body_swap_v1_krea2_example_outdoors.png filter=lfs diff=lfs merge=lfs -text
+images/bfs_body_swap_v1_krea2_example_restaurant.png filter=lfs diff=lfs merge=lfs -text
+images/bfs_body_swap_v1_krea2_example_seated.png filter=lfs diff=lfs merge=lfs -text
+images/bfs_body_swap_v1_krea2_example_yoga.png filter=lfs diff=lfs merge=lfs -text
+images/bfs_qwen_2.1_1.png filter=lfs diff=lfs merge=lfs -text
+images/bfs_qwen_2.1_2.png filter=lfs diff=lfs merge=lfs -text
+images/bfs_qwen_2.1_3.png filter=lfs diff=lfs merge=lfs -text
+images/bfs_qwen_2.1_4.png filter=lfs diff=lfs merge=lfs -text
+images/bfs_qwen_2.1_5.png filter=lfs diff=lfs merge=lfs -text
+images/bfs_qwen_2.1_6.png filter=lfs diff=lfs merge=lfs -text
+images/bfs_qwen_2.1_7.png filter=lfs diff=lfs merge=lfs -text
+images/bfs_qwen_2.1_8.png filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,77 @@
+---
+license: mit
+base_model:
+- Qwen/Qwen-Image-2.1
+- Qwen/Qwen-Image-Edit-2511
+- black-forest-labs/FLUX.2-klein-9B
+- black-forest-labs/FLUX.2-klein-4B
+- black-forest-labs/FLUX.2-klein-base-4B
+- black-forest-labs/FLUX.2-klein-base-9B
+- black-forest-labs/FLUX.2-klein-9b-fp8
+- black-forest-labs/FLUX.2-klein-4b-fp8
+- krea/Krea-2-Raw
+pipeline_tag: image-to-image
+tags:
+- lora
+- qwen-image
+- qwen-image-2.1
+- qwen-image-edit
+- face-swap
+- head-swap
+- body-swap
+- image-editing
+- comfyui
+- bfs
+language:
+- en
+library_name: diffusers
+---
+
+# BFS: Best Face Swap
+
+BFS is a collection of LoRAs for face, head, and body replacement across Qwen Image 2.1, Qwen
+Image Edit 2509/2511, Flux 2 Klein, Krea 2, and experimental LTX-2 workflows. Each release has
+its own base-model, input-order, prompt, and workflow contract. Use the guide for the exact
+version you are loading.
+
+## Guides
+
+- [BFS Head V1 / V1.1 for Qwen Image 2.1](docs/qwen-image-2.1.md)
+- [BFS Body Swap V1.0 for Qwen Image 2.1](docs/qwen-image-2.1-body.md)
+- [Qwen Image Edit 2509/2511](docs/qwen-image-edit.md)
+- [Flux 2 Klein](docs/flux-2-klein.md)
+- [Krea 2](docs/krea-2.md)
+- [Experimental LTX-2 weights](docs/ltx-2.md)
+
+## Downloads at a glance
+
+| Family | Weights | Workflow |
+|---|---|---|
+| Qwen Image 2.1 (head) | [`bfs_head_v1.1_qwen_2.1.safetensors`](bfs_head_v1.1_qwen_2.1.safetensors) (recommended), [`bfs_head_v1.1_alternative_qwen_2.1.safetensors`](bfs_head_v1.1_alternative_qwen_2.1.safetensors), [`bfs_head_v1_qwen_2.1.safetensors`](bfs_head_v1_qwen_2.1.safetensors) |
+| Qwen Image 2.1 (body) | [`bfs_body_swap_v1.0_qwen_2.1.safetensors`](bfs_body_swap_v1.0_qwen_2.1.safetensors) | See the [body swap guide](docs/qwen-image-2.1-body.md) | [`Head Swap V1 Qwen 2.1 Workflow.json`](workflows/Head%20Swap%20V1%20Qwen%202.1%20Workflow.json) |
+| Qwen Image Edit 2509 | [`bfs_face_v1_qwen_image_edit_2509.safetensors`](bfs_face_v1_qwen_image_edit_2509.safetensors), [`bfs_head_v1_qwen_image_edit_2509.safetensors`](bfs_head_v1_qwen_image_edit_2509.safetensors), [`bfs_head_v2_qwen_image_edit_2509.safetensors`](bfs_head_v2_qwen_image_edit_2509.safetensors), [`bfs_head_v3_qwen_image_edit_2509.safetensors`](bfs_head_v3_qwen_image_edit_2509.safetensors), [`bfs_head_v4_qwen_image_edit_2509.safetensors`](bfs_head_v4_qwen_image_edit_2509.safetensors) | See the [Qwen Edit guide](docs/qwen-image-edit.md) |
+| Qwen Image Edit 2511 | [`bfs_head_v5_2511_original.safetensors`](bfs_head_v5_2511_original.safetensors), [`bfs_head_v5_2511_merged_version_rank_16_fp16.safetensors`](bfs_head_v5_2511_merged_version_rank_16_fp16.safetensors), [`bfs_head_v5_2511_merged_version_rank_32_fp32.safetensors`](bfs_head_v5_2511_merged_version_rank_32_fp32.safetensors) | See the [Qwen Edit guide](docs/qwen-image-edit.md) |
+| Flux 2 Klein | [`bfs_head_v1_flux-klein_4b.safetensors`](bfs_head_v1_flux-klein_4b.safetensors), [`bfs_head_v1_flux-klein_9b_step3500_rank128.safetensors`](bfs_head_v1_flux-klein_9b_step3500_rank128.safetensors), [`bfs_head_v1_flux-klein_9b_step3750_rank64.safetensors`](bfs_head_v1_flux-klein_9b_step3750_rank64.safetensors), [`bfs_head_v1.1_optional_flux-klein_4b.safetensors`](bfs_head_v1.1_optional_flux-klein_4b.safetensors) | See the [Flux guide](docs/flux-2-klein.md) |
+| Krea 2 | [`bfs_head_swap_v1.1_krea2.safetensors`](bfs_head_swap_v1.1_krea2.safetensors), [`bfs_body_swap_v1_krea2.safetensors`](bfs_body_swap_v1_krea2.safetensors) | See the [Krea 2 guide](docs/krea-2.md) |
+| Experimental LTX-2 | [`bfs_head_swap_v1_ltx2_ic_lora_5000.safetensors`](bfs_head_swap_v1_ltx2_ic_lora_5000.safetensors), [`bfs_head_swap_v1_ltx2_ic_lora_6000.safetensors`](bfs_head_swap_v1_ltx2_ic_lora_6000.safetensors), [`bfs_head_swap_v1_ltx2_ic_lora_12000_train_3.safetensors`](bfs_head_swap_v1_ltx2_ic_lora_12000_train_3.safetensors) | See the [LTX-2 guide](docs/ltx-2.md) |
+
+## General rules
+
+- Use the LoRA with its matching base model whenever possible.
+- Follow the input order in the version guide; several BFS head releases intentionally use body
+  first and reference face second.
+- Start at LoRA strength `1.0`, then adjust for the base model, resolution, and reference image.
+- The examples and workflows in this repository are demonstrations, not a guarantee of identical
+  results on every image.
+- Do not use or share results involving public figures or people who have not given consent.
+
+## Examples
+
+The existing examples are grouped by workflow family in the individual guides. Qwen Image 2.1
+examples are `images/bfs_qwen_2.1_1.png` through `images/bfs_qwen_2.1_8.png`.
+
+## Support
+
+Feel free to experiment and share your results responsibly. For workflow issues, first update
+ComfyUI and verify that the base model, text encoder, VAE, workflow, and LoRA all belong to the
+same model family.

bfs_body_swap_v1.0_qwen_2.1.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7dc0a53aba4dbc70c204f498936a619d226559da11011f748c389189af46b664
+size 209753576

bfs_body_swap_v1_krea2.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0b3d043714c912c55c525ac68a53f50dbfaa6a024d735c28dfed12cd214a0d79
+size 914160176

bfs_face_v1_qwen_image_edit_2509.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c9f18ba062d0c8aa3c98169005b616bc95866fe9956c2608dcfd29b1e9b20359
+size 590058816

bfs_head_swap_v1.1_krea2.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:abc6c468c578b36e6bd960106ad5886afd3f5321e4fe2de26c05bb5c6585b836
+size 914159816

bfs_head_swap_v1_ltx2_ic_lora_12000_train_3.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:27bb754d5c5029e64131d854217ddb37dfa17eb07864201e911e9c89e633ecaf
+size 1308756368

bfs_head_swap_v1_ltx2_ic_lora_5000.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e11d18bb009b2c2a938b99a32f3ab140705fdbea7b65029c3cda3ceeb89e654c
+size 1308756368

bfs_head_swap_v1_ltx2_ic_lora_6000.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4021b4143bd935e565aeaaf5bf8ef43e8f3aa3faf7a9ae0a73f1879342f32d12
+size 1308756368

bfs_head_v1.1_alternative_qwen_2.1.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c5332bbc2f826856e7a09bce9740e97d39514b849368212e1f2c3c2c2d81c217
+size 318821008

bfs_head_v1.1_optional_flux-klein_4b.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8234563b48b66b22aaaa0df9b4c70636e65f06d443c3b20a9356425040255839
+size 1478513544

bfs_head_v1.1_qwen_2.1.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d1d748d5601077f3b6d05766f6823510e901970d916afa404a97e85dc92fa88e
+size 260096144

bfs_head_v1_flux-klein_4b.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:77aabcd6144612dd4873080b99283dd9a27977712548bee161894f62c91ff09c
+size 369644384

bfs_head_v1_flux-klein_9b_step3500_rank128.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:70d8aaf332d710b905d5085afaa87c3ef577edffd54ffcfadeb8c47a854f9044
+size 662729912

bfs_head_v1_flux-klein_9b_step3750_rank64.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:49907a240602488d2846861d7570dd95d80609f44be7e3347965735712b1be16
+size 331379568

bfs_head_v1_qwen_2.1.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:facf18a8c4c2f62578c034fc4e7cbdb964c63fb2d678fe0d3d41677fba3d0606
+size 318820976

bfs_head_v1_qwen_image_edit_2509.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d71ee1612e4b416fc0f62a1a59a3d66b7ee6d86a1a7c35aacc2f929e0326f68d
+size 590058808

bfs_head_v2_qwen_image_edit_2509.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a928f106fcba6e4b196ef511289f075b4690eba0f6bc3fa3755fdba2ec2a428c
+size 147687224

bfs_head_v3_qwen_image_edit_2509.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:49aac5a55175ab5ebd89ffdae4c336eaca1899f5b725b70c0c039f6d4a8d86e6
+size 73957904

bfs_head_v4_qwen_image_edit_2509.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd3278c33c7df36bb87169d0a283851d60e88ecac48123b5c8dcc82551d6e402
+size 147794944

bfs_head_v5_2511_merged_version_rank_16_fp16.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1315a08947e5d6d7c53ea4fc59f272e0d54efd7f2da35999b7d873a7cf4fa89b
+size 307464216

bfs_head_v5_2511_merged_version_rank_32_fp32.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0a137e61245781412421f5dee5db4ccac28b6c9952c042d1123a84609107cd10
+size 1206402600

bfs_head_v5_2511_original.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5055ec271228d4767fd327efa59dbdea6f2a628d6d890101eb4fd5015cb1ac57
+size 147685552

docs/flux-2-klein.md

@@ -0,0 +1,33 @@
+# Flux 2 Klein Head-Swap LoRAs
+
+These BFS weights target Flux 2 Klein 4B/9B variants. They use the body-first input contract:
+
+- **Image 1:** target body/base image
+- **Image 2:** reference face/head
+
+Do not reverse the images for these releases.
+
+## Downloads
+
+- [`bfs_head_v1_flux-klein_4b.safetensors`](../bfs_head_v1_flux-klein_4b.safetensors)
+- [`bfs_head_v1_flux-klein_9b_step3500_rank128.safetensors`](../bfs_head_v1_flux-klein_9b_step3500_rank128.safetensors)
+- [`bfs_head_v1_flux-klein_9b_step3750_rank64.safetensors`](../bfs_head_v1_flux-klein_9b_step3750_rank64.safetensors)
+- [`bfs_head_v1.1_optional_flux-klein_4b.safetensors`](../bfs_head_v1.1_optional_flux-klein_4b.safetensors)
+
+## Prompt
+
+```text
+head_swap: start with Picture 1 as the base image, keeping its lighting, environment, and background. remove the head from Picture 1 completely and replace it with the head from Picture 2, strictly preserving the hair, eye color, and nose structure of Picture 2. copy the direction of the eye, head rotation, and micro-expressions from Picture 1. Keep the body, clothing, pose, framing, and background intact.
+```
+
+You can add a short expression or pose instruction when it is visible in the target image. Keep
+the edit focused on the head and avoid requesting an unrelated scene change in the same prompt.
+
+## Workflow
+
+[`workflows/Head Swap V1 Flux 2 Klein 4b_9b (base_distill).json`](../workflows/Head%20Swap%20V1%20Flux%202%20Klein%204b_9b%20(base_distill).json)
+
+Start around LoRA strength `1.0`; adjust after checking identity preservation, hair boundaries,
+expression transfer, and skin blending.
+
+[Back to the repository index](../README.md)

docs/krea-2.md

@@ -0,0 +1,48 @@
+# Krea 2 Head and Body Swap LoRAs
+
+## Head Swap V1.1
+
+[`bfs_head_swap_v1.1_krea2.safetensors`](../bfs_head_swap_v1.1_krea2.safetensors) is the updated
+Krea 2 head-swap release and replaces the earlier V1 file.
+
+Use the included [`Head Swap Krea 2 - V1 Simple Workflow.json`](../workflows/Head%20Swap%20Krea%202%20-%20V1%20Simple%20Workflow.json).
+
+Inputs:
+
+- Image 1: base image
+- Image 2: reference head
+
+Prompt:
+
+```text
+head_swap: replace the head with the reference head.
+```
+
+## Body Swap V1 (experimental)
+
+[`bfs_body_swap_v1_krea2.safetensors`](../bfs_body_swap_v1_krea2.safetensors) replaces the
+source person with the reference person. Exact pose transfer is not guaranteed.
+
+Use the included [`Body Swap Krea 2 - V1 Simple Workflow.json`](../workflows/Body%20Swap%20Krea%202%20-%20V1%20Simple%20Workflow.json).
+
+Inputs:
+
+- Image 1: base image
+- Image 2: reference person
+
+Prompt:
+
+```text
+body_swap: replace the person with the reference person.
+```
+
+To request a new background, append:
+
+```text
+BG: [describe the new background]
+```
+
+Start around LoRA strength `0.5` for body swap and adjust from there. The head workflow can be
+combined with the body workflow when additional head-identity control is useful.
+
+[Back to the repository index](../README.md)

docs/ltx-2.md

@@ -0,0 +1,16 @@
+# Experimental LTX-2 Head-Swap Weights
+
+The repository also contains experimental LTX-2 in-context LoRA checkpoints:
+
+- [`bfs_head_swap_v1_ltx2_ic_lora_5000.safetensors`](../bfs_head_swap_v1_ltx2_ic_lora_5000.safetensors)
+- [`bfs_head_swap_v1_ltx2_ic_lora_6000.safetensors`](../bfs_head_swap_v1_ltx2_ic_lora_6000.safetensors)
+- [`bfs_head_swap_v1_ltx2_ic_lora_12000_train_3.safetensors`](../bfs_head_swap_v1_ltx2_ic_lora_12000_train_3.safetensors)
+
+These files are experimental and are kept separate from the Qwen, Flux, and Krea release guides.
+Use them only with an LTX-2 workflow that supports the corresponding in-context LoRA contract.
+No LTX-2 workflow is bundled in this repository at the moment.
+
+Do not assume that the Qwen or Flux image input order applies to LTX-2. Follow the input contract
+of the LTX-2 workflow you are using and test the checkpoint before a long generation.
+
+[Back to the repository index](../README.md)

docs/qwen-image-2.1-body.md

@@ -0,0 +1,71 @@
+# BFS Body Swap V1.0 for Qwen Image 2.1
+
+[`bfs_body_swap_v1.0_qwen_2.1.safetensors`](../bfs_body_swap_v1.0_qwen_2.1.safetensors) replaces
+the **whole person** in a scene with the person from a reference image, keeping the scene's pose,
+framing, lighting and background. It is separate from the head-swap weights: use this one when the
+body, clothing and proportions should come from the reference, not only the head.
+
+Base model: [`Qwen/Qwen-Image-2.1`](https://huggingface.co/Qwen/Qwen-Image-2.1). Strength `1.0`.
+
+## Inputs
+
+- **Image 1 / `<image1>`:** the scene. Its pose, framing, lighting and background are kept.
+- **Image 2 / `<image2>`:** the reference person, whose face, clothing and body proportions are
+  transferred.
+
+Keep this order.
+
+## Prompt
+
+```text
+body_swap: start with <image1> as the base image, keeping its lighting, environment, and background. replace the body from <image1> with the body from <image2>, strictly preserving the clothing, body shape and proportions from <image2>. copy the pose, direction of the eye, head rotation, micro expressions from <image1>
+```
+
+## The reference image matters more than anything else
+
+This model was trained with references in one consistent format: **the person standing, full body,
+facing the camera, on a plain light background**, filling most of a square frame. A reference in
+that format is inside the distribution the model learned. A tight portrait with a busy background
+is not, and that is where results fall apart: the reference's own background starts bleeding into
+the output and identity gets weaker.
+
+Two things follow from that.
+
+**Give the reference room, not resolution.** If you feed both images at their native size, a
+reference larger than the scene takes more tokens and starts overriding it. Scaling the reference
+to roughly 0.59 MP (about 768x768 of area) while keeping its aspect ratio measurably improved how
+well the scene's background survived, at every output size tested. Scaling the reference *up* to
+match the scene's size was the worst setting measured, worse than doing nothing.
+
+**Normalize the reference first if it is not full body.** A single pass through Qwen Image 2.1
+itself, with no LoRA, turns a portrait into a reference in the trained format:
+
+```text
+full body studio reference photo of the person from <image1>: standing upright on both feet, facing the camera, front view, camera at chest height and perfectly level, arms relaxed at the sides, neutral expression, looking straight at the camera. plain seamless light background, soft even studio lighting, no props and no cast shadows on the backdrop. the person fills almost the whole height of the square frame, from the top of the head down to the shoes, with a small even margin above and below. strictly keep the same face, hairstyle, clothing and body proportions as <image1>. sharp photographic detail
+```
+
+Run that at 1024x1024 and use the result as `<image2>`. It reconstructs legs and shoes that the
+original crop did not show, so expect some loss of facial likeness in the reference itself; whether
+the round trip is worth it depends on how far your source is from the trained format.
+
+## ComfyUI setup
+
+1. Put the file in `ComfyUI/models/loras/`.
+2. Load the Qwen Image 2.1 diffusion model, the Qwen3-VL text encoder and the matching VAE.
+3. Scale the **scene** to about 2 MP and the **reference** to about 0.59 MP, both preserving aspect
+   ratio (`ImageScaleToTotalPixels` does this).
+4. Feed both to `TextEncodeQwenImage21` with `resolution: 0`, scene first.
+5. Take the sampler's latent from that node's `latent` output: it is already sized to the scene, so
+   the aspect ratio follows without a separate `EmptyLatentImage` to keep in sync.
+
+## Limitations
+
+Faces get less accurate as the subject gets smaller in frame; the drop is measurable and it is the
+main open weakness. Very large changes between scene and reference (different sex, age or build at
+the same time) are the cases most likely to fail outright. Pose is followed approximately, not
+exactly, and prompt wording moves it only a few percent. Always check the result for unintended
+changes to the background, clothing and text.
+
+Do not use or share results involving public figures or people who have not given consent.
+
+[Back to the repository index](../README.md)

docs/qwen-image-2.1.md

@@ -0,0 +1,134 @@
+# BFS Head V1 / V1.1 for Qwen Image 2.1
+
+The BFS head-swap LoRA for [`Qwen/Qwen-Image-2.1`](https://huggingface.co/Qwen/Qwen-Image-2.1)
+ships in three versions:
+[`bfs_head_v1.1_qwen_2.1.safetensors`](../bfs_head_v1.1_qwen_2.1.safetensors) (recommended),
+[`bfs_head_v1.1_alternative_qwen_2.1.safetensors`](../bfs_head_v1.1_alternative_qwen_2.1.safetensors),
+and the original [`bfs_head_v1_qwen_2.1.safetensors`](../bfs_head_v1_qwen_2.1.safetensors).
+All three are drop-in replacements for each other: same workflow, same prompt, strength `1.0`.
+
+This release is separate from the Qwen Image Edit 2509/2511 weights. Use the Qwen Image 2.1
+base model and the included workflow:
+
+[`workflows/Head Swap V1 Qwen 2.1 Workflow.json`](../workflows/Head%20Swap%20V1%20Qwen%202.1%20Workflow.json)
+
+## Versions
+
+### V1.1: reduced skin softening
+
+V1 transfers identity well, but it also softens skin: pores, freckles and fine lines get smoothed
+out, and results can read as slightly plastic. That softening does not come from the whole LoRA.
+It comes from one specific part of it.
+
+The LoRA touches six modules in each of the 32 transformer blocks: four attention projections
+(`to_q`, `to_k`, `to_v`, `to_out.0`) and two MLP ones (`img_mlp.gate_up`, `img_mlp.out`).
+Disabling them group by group and comparing generations under a fixed seed, prompt and sampler
+shows the blur is carried by the `img_mlp.gate_up` path. Attention routes information from the
+reference head; the MLP path is where the model's notion of how a surface looks gets rewritten,
+so that is where training pulled skin texture toward the dataset.
+
+V1.1 is V1 with the `img_mlp.gate_up` modules removed from the second half of the network
+(blocks 16-31), 16 of the 192 modules. Every attention module, every `img_mlp.out` module and
+the first-half `gate_up` modules are byte-identical to V1. Nothing was rescaled, re-ranked or
+retrained.
+
+The cut stops at block 16 for a reason. The early-block MLP path is what makes the LoRA copy head
+rotation, eye direction and expression from `<image1>`. Removing `img_mlp.gate_up` everywhere
+yields even more skin detail, but the LoRA then stops following the source pose and degenerates
+into pasting the reference head onto the body, which is close to what the base model already does
+without any LoRA. V1.1 sits at the point where texture improves while that behaviour still works.
+
+Measured across five image pairs at a fixed seed, V1.1 raises skin micro-detail inside the face
+(bilateral residual) by about a third over V1, closing roughly two thirds of the distance to what
+the base model produces without any LoRA. Identity similarity and head-pose error stay close to
+V1. Expression transfer still works, but it is weaker than V1. On a deliberately extreme test
+case (a wide-open mouth in `<image1>`), V1 reproduces the expression more fully than V1.1.
+
+### V1.1 alternative - stronger expression transfer
+
+`bfs_head_v1.1_alternative_qwen_2.1.safetensors` attacks the same problem from the other side.
+Instead of removing `img_mlp.gate_up` modules, it keeps all 192 modules and scales the whole
+`img_mlp` path down to 35% of its trained magnitude, leaving the 128 attention modules untouched.
+Scaling preserves the direction of the weight update and only reduces its strength, so the
+behaviour that depends on the MLP path survives instead of being switched off.
+
+In testing this version keeps more skin detail than V1.1 and copies expression from `<image1>`
+noticeably better - on the wide-open-mouth test case it reproduces the expression more fully than
+V1.1 and more fully than V1 itself. It pays for that with a slightly larger head-pose error, about
+one degree on average across the test pairs.
+
+Pick this one when the source image has a strong expression or the mouth and eyes carry the shot.
+Pick V1.1 when head angle fidelity matters most.
+
+
+### V1: original release
+
+The original release, kept for reproducibility and for anyone who already tuned a pipeline around
+it. It has the softest skin of the three. Note that the V1.1 alternative above matches or beats it
+at copying expression while keeping more skin detail, so there is little reason to prefer V1 unless
+you want to handle texture yourself in post.
+
+
+## Inputs
+
+The workflow uses two images:
+
+- **Image 1 / `<image1>`:** the target or base image. Its body, pose, composition, lighting, and
+  environment provide the foundation for the result.
+- **Image 2 / `<image2>`:** the reference face/head. Its identity, hair, eye color, nose
+  structure, and other visible head features are transferred.
+
+Keep this order. Reversing the images changes which person is treated as the target.
+
+## Prompt
+
+The included workflow uses this starting prompt:
+
+```text
+head_swap: start with <image1> as the base image, keeping its lighting, environment, and background. remove the head from <image1> completely and replace it with the head from <image2>, strictly preserving the hair, eye color, nose structure from <image2>. copy the direction of the eye, head rotation, micro expressions from <image1>, high quality, sharp details, 4k
+```
+
+Replace `<image1>` and `<image2>` only if your node or workflow uses a different reference-token
+notation. Keep the edit localized to the head when the body and background should remain unchanged.
+
+## ComfyUI setup
+
+1. Put `bfs_head_v1.1_qwen_2.1.safetensors` (or one of the other two versions) in `ComfyUI/models/loras/`.
+2. Load the Qwen Image 2.1 diffusion model, Qwen3-VL text encoder, and matching Qwen Image 2.1
+   VAE required by the workflow.
+3. Open the included workflow and connect the target image to Image 1 and the reference head to
+   Image 2.
+4. Start with LoRA strength `1.0` and adjust only after checking the identity, hair boundary,
+   skin blending, and expression transfer.
+
+The workflow also contains an optional Pruna 8-step acceleration LoRA. It is not part of BFS and
+can be disabled if it is not installed.
+
+## Resolution and quality
+
+Qwen Image 2.1 supports native high-resolution generation. The included workflow starts from a
+moderate resolution and supports custom sizing in multiples of 32. For this head-swap use case,
+larger output sizes can help retain facial detail and reduce the broad color or softness shifts
+that may appear at lower resolutions. Higher resolution does not guarantee perfect preservation.
+
+## Examples
+
+<p align="center">
+  <img src="../images/bfs_qwen_2.1_1.png" width="49%" />
+  <img src="../images/bfs_qwen_2.1_2.png" width="49%" />
+  <img src="../images/bfs_qwen_2.1_3.png" width="49%" />
+  <img src="../images/bfs_qwen_2.1_4.png" width="49%" />
+  <img src="../images/bfs_qwen_2.1_5.png" width="49%" />
+  <img src="../images/bfs_qwen_2.1_6.png" width="49%" />
+  <img src="../images/bfs_qwen_2.1_7.png" width="49%" />
+  <img src="../images/bfs_qwen_2.1_8.png" width="49%" />
+</p>
+
+## Limitations and responsible use
+
+Small faces, extreme angles, occlusion, hands over the face, hair boundaries, and strong lighting
+changes can reduce fidelity. Always inspect the result for unintended changes to the body,
+clothing, background, text, and identity. Do not use or share results involving public figures or
+people who have not given consent.
+
+[Back to the repository index](../README.md)

docs/qwen-image-edit.md

@@ -0,0 +1,67 @@
+# Qwen Image Edit LoRAs
+
+This family contains the original BFS releases for Qwen Image Edit 2509 and the later 2511
+head-swap continuation. Use the matching Qwen base model and keep the input order for the selected
+version.
+
+## Downloads
+
+### Qwen Image Edit 2509
+
+- [`bfs_face_v1_qwen_image_edit_2509.safetensors`](../bfs_face_v1_qwen_image_edit_2509.safetensors)
+- [`bfs_head_v1_qwen_image_edit_2509.safetensors`](../bfs_head_v1_qwen_image_edit_2509.safetensors)
+- [`bfs_head_v2_qwen_image_edit_2509.safetensors`](../bfs_head_v2_qwen_image_edit_2509.safetensors)
+- [`bfs_head_v3_qwen_image_edit_2509.safetensors`](../bfs_head_v3_qwen_image_edit_2509.safetensors)
+- [`bfs_head_v4_qwen_image_edit_2509.safetensors`](../bfs_head_v4_qwen_image_edit_2509.safetensors)
+
+### Qwen Image Edit 2511
+
+- [`bfs_head_v5_2511_original.safetensors`](../bfs_head_v5_2511_original.safetensors)
+- [`bfs_head_v5_2511_merged_version_rank_16_fp16.safetensors`](../bfs_head_v5_2511_merged_version_rank_16_fp16.safetensors)
+- [`bfs_head_v5_2511_merged_version_rank_32_fp32.safetensors`](../bfs_head_v5_2511_merged_version_rank_32_fp32.safetensors)
+
+The V5 files are the latest 2511 release. The merged versions combine the V5 2511 training with
+the earlier V4 behavior; the FP16 file is the practical default, while the FP32 file is kept for
+maximum numerical fidelity when storage and VRAM allow it.
+
+## Input order
+
+| Version | Image 1 | Image 2 |
+|---|---|---|
+| Face V1 | Reference face | Target body |
+| Head V1/V2 | Reference face | Target body |
+| Head V3/V4/V5 | Target body | Reference face |
+
+The V3, V4, and V5 releases intentionally use the inverted body-first order. Do not reuse the
+Face V1/V2 order with those files.
+
+## Prompts
+
+Face V1:
+
+```text
+face swap face from Image 1 to Image 2. swap only the face, match the skin tone, and keep the target pose and lighting.
+```
+
+Head V1/V2:
+
+```text
+head swap from Image 1 to Image 2
+```
+
+Head V3/V4/V5:
+
+```text
+head_swap: start with Picture 1 as the base image, keeping its lighting, environment, and background. remove the head from Picture 1 completely and replace it with the head from Picture 2, strictly preserving the hair, eye color, and nose structure of Picture 2. copy the eye direction, head rotation, and micro-expressions from Picture 1.
+```
+
+## Workflow
+
+The repository contains workflows for the V3, V4, and V5-style Qwen setup. Use the workflow that
+matches the base model and version, then start around LoRA strength `1.0` and adjust based on the
+reference image and resolution.
+
+This family is separate from the [Qwen Image 2.1 release](qwen-image-2.1.md), which uses
+`<image1>`/`<image2>` in its dedicated workflow.
+
+[Back to the repository index](../README.md)

hf_remote_info.json

@@ -0,0 +1,251 @@
+{
+  "id": "Alissonerdx/BFS-Best-Face-Swap",
+  "author": "Alissonerdx",
+  "sha": "c758d55502463bc5eb281326c107d0850d7a8b4a",
+  "created_at": "2025-11-07T09:50:28+00:00",
+  "last_modified": "2026-08-10T19:53:10+00:00",
+  "private": false,
+  "disabled": false,
+  "downloads": 110991,
+  "gated": false,
+  "inference": "warm",
+  "likes": 757,
+  "library_name": "diffusers",
+  "tags": [
+    "diffusers",
+    "lora",
+    "qwen-image-edit",
+    "face-swap",
+    "head-swap",
+    "body-swap",
+    "image-editing",
+    "ai-generated",
+    "comfyui",
+    "bfs",
+    "flux 2",
+    "flux",
+    "klein",
+    "krea2",
+    "image-to-image",
+    "en",
+    "base_model:Qwen/Qwen-Image-Edit-2511",
+    "base_model:adapter:Qwen/Qwen-Image-Edit-2511",
+    "license:mit",
+    "region:us"
+  ],
+  "pipeline_tag": "image-to-image",
+  "card_data": {},
+  "config": {},
+  "siblings": [
+    {
+      "rfilename": ".gitattributes"
+    },
+    {
+      "rfilename": "README.md"
+    },
+    {
+      "rfilename": "bfs_body_swap_v1_krea2.safetensors"
+    },
+    {
+      "rfilename": "bfs_face_v1_qwen_image_edit_2509.safetensors"
+    },
+    {
+      "rfilename": "bfs_head_swap_v1.1_krea2.safetensors"
+    },
+    {
+      "rfilename": "bfs_head_swap_v1_ltx2_ic_lora_12000_train_3.safetensors"
+    },
+    {
+      "rfilename": "bfs_head_swap_v1_ltx2_ic_lora_5000.safetensors"
+    },
+    {
+      "rfilename": "bfs_head_swap_v1_ltx2_ic_lora_6000.safetensors"
+    },
+    {
+      "rfilename": "bfs_head_v1.1_optional_flux-klein_4b.safetensors"
+    },
+    {
+      "rfilename": "bfs_head_v1_flux-klein_4b.safetensors"
+    },
+    {
+      "rfilename": "bfs_head_v1_flux-klein_9b_step3500_rank128.safetensors"
+    },
+    {
+      "rfilename": "bfs_head_v1_flux-klein_9b_step3750_rank64.safetensors"
+    },
+    {
+      "rfilename": "bfs_head_v1_qwen_image_edit_2509.safetensors"
+    },
+    {
+      "rfilename": "bfs_head_v2_qwen_image_edit_2509.safetensors"
+    },
+    {
+      "rfilename": "bfs_head_v3_qwen_image_edit_2509.safetensors"
+    },
+    {
+      "rfilename": "bfs_head_v4_qwen_image_edit_2509.safetensors"
+    },
+    {
+      "rfilename": "bfs_head_v5_2511_merged_version_rank_16_fp16.safetensors"
+    },
+    {
+      "rfilename": "bfs_head_v5_2511_merged_version_rank_32_fp32.safetensors"
+    },
+    {
+      "rfilename": "bfs_head_v5_2511_original.safetensors"
+    },
+    {
+      "rfilename": "images/ComfyUI_04745_.png"
+    },
+    {
+      "rfilename": "images/ComfyUI_04755_.png"
+    },
+    {
+      "rfilename": "images/ComfyUI_04759_.png"
+    },
+    {
+      "rfilename": "images/ComfyUI_04761_.png"
+    },
+    {
+      "rfilename": "images/bfs_body_swap_v1_krea2_example_closeup.png"
+    },
+    {
+      "rfilename": "images/bfs_body_swap_v1_krea2_example_kitchen.png"
+    },
+    {
+      "rfilename": "images/bfs_body_swap_v1_krea2_example_outdoors.png"
+    },
+    {
+      "rfilename": "images/bfs_body_swap_v1_krea2_example_restaurant.png"
+    },
+    {
+      "rfilename": "images/bfs_body_swap_v1_krea2_example_seated.png"
+    },
+    {
+      "rfilename": "images/bfs_body_swap_v1_krea2_example_yoga.png"
+    },
+    {
+      "rfilename": "workflows/Body Swap Krea 2 - V1 Simple Workflow.json"
+    },
+    {
+      "rfilename": "workflows/Head Swap 2511 - V5 Simple Workflow.json"
+    },
+    {
+      "rfilename": "workflows/Head Swap Krea 2 - V1 Simple Workflow.json"
+    },
+    {
+      "rfilename": "workflows/Head Swap V1 Flux 2 Klein 4b_9b (base_distill).json"
+    },
+    {
+      "rfilename": "workflows/Head Swap V3 Simple Workflow (With Lightining LoRA) .json"
+    },
+    {
+      "rfilename": "workflows/Head Swap V3 Simple Workflow (Without Lightining LoRA).json"
+    },
+    {
+      "rfilename": "workflows/Head Swap V4 Simple Workflow.json"
+    }
+  ],
+  "spaces": [
+    "cruisewagner2220/Qwen-Image-Edit-Rapid-AIO-Loras-Experimental-neo",
+    "M3st3rJ4k3l/FLUX.2-Klein-Multi-LoRA",
+    "linoyts/Flux2-Klein-Face-Swap",
+    "ovi054/Qwen-Image-Edit-2511-LoRA",
+    "laruss5/Flux2-Klein-Face-Swap",
+    "pnemrow/Qwen-Image-Edit-Rapid-AIO-Loras-Experimental-neo",
+    "Mccscs2/Qwen-Image-Edit-Rapid-AIO-Loras-Experimental",
+    "DanFrmSpace/BFS-Best-Face-Swap",
+    "mmmat32/FLUX.2-Klein-Multi-LoRA-3D-Angle",
+    "jt65/Qwen-Image-Edit-2511-LoRA-nsfw",
+    "prithivMLmods/FLUX.2-Klein-LoRA-Studio",
+    "someone-in-the-world/Selectable-Face-Swap",
+    "GTO83/Qwen-Image-Edit-Rapid-AIO-Loras-Experimental-neo",
+    "aet256/Qwen-Image-Edit-Rapid-AIO-Loras-Experimental",
+    "Onise/FLUX.2-Klein-LoRA-Studio",
+    "ahuggingface01/BFS-Best-Face-Swap",
+    "linoyts/Face-Swap-Frame",
+    "e1357/FLUX.2-Klein-Multi-LoRA",
+    "Alissonerdx/BFS-Best-Face-Swap",
+    "M3st3rJ4k3l/Qwen-Image-Edit-Rapid-AIO-Loras-Experimental",
+    "anonyketaa/Alissonerdx-BFS-Best-Face-Swap",
+    "ishrak-8495/Alissonerdx-BFS-Best-Face-Swap",
+    "Furkangundog18/Flux2-Klein-Face-Swap",
+    "Sebastiankay/bildbearbeitung",
+    "idcuso/Alissonerdx-BFS-Best-Face-Swap",
+    "blablaai11/Alissonerdx-BFS-Best-Face-Swap",
+    "linoyts/Subject-Swap-Frame",
+    "cereeenn120/Hogwarts_SBox_AI-RPG",
+    "yayite/FLUX.2-Klein-Multi-LoRA",
+    "Briante/BFS-Best-Face-Swap",
+    "Revo420/Alissonerdx-BFS-Best-Face-Swap",
+    "ideryang/Alissonerdx-BFS-Best-Face-Swap",
+    "ideryang/BFS-Best-Face-Swap",
+    "Mrwigglezz/Alissonerdx-BFS-Best-Face-Swap",
+    "ikolomiets/Flux2-Klein-Face-Swap",
+    "Dmitry1313/facefusion-api",
+    "akashgundgire78/zero123-views",
+    "raubatz/Qwen-Image-Edit-Rapid-AIO-Loras-Experimental-neo",
+    "masterp3000/FLUX.2-Klein-Multi-LoRA",
+    "MaxBardi/FLUX.2-Klein-LoRA-Studio",
+    "BigFaceChase/Flux2-Klein-Face-Swap",
+    "Hittler-AI/FLUX.2-Klein-Multi-LoRA",
+    "aniketb04/Alissonerdx-BFS-Best-Face-Swap",
+    "Pippolononone/Alissonerdx-BFS-Best-Face-Swap",
+    "Dnidbdnx/Alissonerdx-BFS-Best-Face-Swap",
+    "pyo8777/Alissonerdx-BFS-Best-Face-Swap",
+    "Jobanpreet/faceswap",
+    "Dreamer333/Alissonerdx-BFS-Best-Face-Swap",
+    "Idlyex/BFS-Best-Face-Swap",
+    "yuhuiwang/Alissonerdx-BFS-Best-Face-Swap",
+    "Andrew115/Alissonerdx-BFS-Best-Face-Swap",
+    "pormungtai/QwenPost-Image-Edit-Rapid-AIO-Loras",
+    "00Boobs00/Qwen-Image-Edit-Rapid-AIO-Lorasses",
+    "Arafat602/Alissonerdx-BFS-Best-Face-Swap",
+    "ziffir/Face-swap-FLUX-2",
+    "GioVar/Qwen-Image-Edit-Rapid-AIO-Loras-Experimental",
+    "DarkJ/Alissonerdx-BFS-Best-Face-Swap",
+    "chokoarbejde/Qwen-Image-Edit-Rapid-AIO-Loras",
+    "Rileygsjsj828292/Alissonerdx-BFS-Best-Face-Swap",
+    "priyesh08/pn-BFS-Best-Face-Swap",
+    "cereeenn120/AI-RPG",
+    "Karter93/QwenPost-Image-Edit-Rapid-AIO-Loras",
+    "noumanvyom/Alissonerdx-BFS-Best-Face-Swap",
+    "mr2along/Bfs",
+    "mr2along/Flux2-Klein",
+    "NYBinSJ/Alissonerdx-BFS-Best-Face-Swap",
+    "deepvoice/Alissonerdx-BFS-Best-Face-Swap",
+    "Vaadadim/Alissonerdx-BFS-Best-Face-Swap",
+    "airpm21/Alissonerdx-BFS-Best-Face-Swap",
+    "sinhachetu/Flux2-Klein-Face-Swap",
+    "rothel888/Alissonerdx-BFS-Best-Face-Swap",
+    "mcmcowen/Flux2-Klein-Face-Swap",
+    "GeBibawy/BFS-Best-Face-Swap",
+    "leonelhs/Flux2-Klein-Face-Swap",
+    "ilian007/FaceSwap",
+    "Poppy222/Alissonerdx-BFS-Best-Face-Swap",
+    "GiorgioV/FLUX.2-Klein-LoRA-Studio12",
+    "dxxslol/Alissonerdx-BFS-Best-Face-Swap",
+    "lahlaliy/Alissonerdx-BFS-Best-Face-Swap",
+    "Micha0103/Alissonerdx-BFS-Best-Face-Swap",
+    "ziffir/Flux2-Klein-Face-Swap",
+    "arabago96/FLUX.2-Klein-LoRA-Studio",
+    "Juliozama1989/FLUX.2-Klein-LoRA-Studio",
+    "Burberman/Alissonerdx-BFS-Best-Face-Swap",
+    "ad242/Flux2-Klein-Face-Swap",
+    "HHClub/Flux2-Klein-Face-Swap",
+    "grnacho/Flux2-Klein-Face-Swap",
+    "ranobe0/Flux2-Klein-Face-Swap2",
+    "EVO817/Flux2-Klein-Face-Swap",
+    "dssddsdf/flux-klein-studio",
+    "Oiknoohs/Alissonerdx-BFS-Best-Face-Swap",
+    "Fernk/Alissonerdx-BFS-Best-Face-Swap",
+    "qu33fy/Alissonerdx-BFS-Best-Face-Swap",
+    "Karter93/FLUX.2-Klein-LoRA-Studio12",
+    "Jonnyquang104/Flux2-Klein-Face-Swap",
+    "Tshovon/Flux2-Klein-Face-Swap",
+    "Jaxxtherapper/Flux2-Klein-Face-Swap",
+    "ofunrein/Alissonerdx-BFS-Best-Face-Swap",
+    "EldMans/Flux2-Klein-Face-Swap",
+    "jp123454321/Flux2-Klein-Face-Swap"
+  ]
+}

hf_upload.err.log

@@ -0,0 +1,594 @@
+Start hashing 11 files.
+Finished hashing 11 files.
+
+
+
+
  ...ap_v1.1_krea2.safetensors:   5%|4         | 41.9MB /  914MB            
+
+
  ...ap_v1.1_krea2.safetensors:   5%|4         | 41.9MB /  914MB            
+
+
+
  ...swap_v1_krea2.safetensors:  12%|#1        |  109MB /  914MB            
+
+
  ...ap_v1.1_krea2.safetensors:  17%|#7        |  159MB /  914MB            
+
+
+
  ...swap_v1_krea2.safetensors:  12%|#1        |  109MB /  914MB            
+
+
  ...ap_v1.1_krea2.safetensors:  30%|###       |  277MB /  914MB            
+
+
+
  ...swap_v1_krea2.safetensors:  24%|##3       |  218MB /  914MB            
+
+
  ...ap_v1.1_krea2.safetensors:  42%|####2     |  386MB /  914MB            
+
+
+
  ...swap_v1_krea2.safetensors:  37%|###6      |  335MB /  914MB            
+
+
  ...ap_v1.1_krea2.safetensors:  55%|#####5    |  503MB /  914MB            
+
+
+
  ...swap_v1_krea2.safetensors:  50%|####9     |  453MB /  914MB            
+
+
  ...ap_v1.1_krea2.safetensors:  69%|######8   |  629MB /  914MB            
+
+
+
  ...swap_v1_krea2.safetensors:  62%|######2   |  570MB /  914MB            
+
+
  ...ap_v1.1_krea2.safetensors:  82%|########1 |  747MB /  914MB            
+
+
+
  ...swap_v1_krea2.safetensors:  76%|#######6  |  696MB /  914MB            
+
+
  ...ap_v1.1_krea2.safetensors:  95%|#########4|  864MB /  914MB            
+
+
+
  ...swap_v1_krea2.safetensors:  89%|########9 |  814MB /  914MB            
+
+
  ...ap_v1.1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
+
  ...swap_v1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
  ...ap_v1.1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
+
  ...swap_v1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
+
+
  ...krea2_example_closeup.png:   3%|3         | 86.0kB / 2.66MB            
+
+
+
+
+
  ...krea2_example_kitchen.png:   3%|3         | 90.0kB / 2.79MB            
+
+
+
+
+
+
  ...rea2_example_outdoors.png:   3%|3         | 82.2kB / 2.54MB            
+
+
+
+
+
+
+
  ...a2_example_restaurant.png:   3%|3         |  102kB / 3.16MB            
+
+
+
+
+
+
+
+
  ..._krea2_example_seated.png:   3%|3         | 80.6kB / 2.50MB            
+
+
+
+
+
+
+
+
+
  ...v1_krea2_example_yoga.png:   3%|3         | 87.2kB / 2.70MB            
+
+
  ...ap_v1.1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
+
  ...swap_v1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
+
+
  ...krea2_example_closeup.png:   3%|3         | 86.0kB / 2.66MB            
+
+
+
+
+
  ...krea2_example_kitchen.png:   3%|3         | 90.0kB / 2.79MB            
+
+
+
+
+
+
  ...rea2_example_outdoors.png:   3%|3         | 82.2kB / 2.54MB            
+
+
+
+
+
+
+
  ...a2_example_restaurant.png:   3%|3         |  102kB / 3.16MB            
+
+
+
+
+
+
+
+
  ..._krea2_example_seated.png:   3%|3         | 80.6kB / 2.50MB            
+
+
+
+
+
+
+
+
+
  ...v1_krea2_example_yoga.png:   3%|3         | 87.2kB / 2.70MB            
+
+
+
  ...ap_v1.1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
+
  ...swap_v1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
+
+
  ...krea2_example_closeup.png:   3%|3         | 86.0kB / 2.66MB            
+
+
+
+
+
  ...krea2_example_kitchen.png:   3%|3         | 90.0kB / 2.79MB            
+
+
+
+
+
+
  ...rea2_example_outdoors.png:   3%|3         | 82.2kB / 2.54MB            
+
+
+
+
+
+
+
  ...a2_example_restaurant.png:   3%|3         |  102kB / 3.16MB            
+
+
+
+
+
+
+
+
  ..._krea2_example_seated.png:   3%|3         | 80.6kB / 2.50MB            
+
+
+
+
+
+
+
+
+
  ...v1_krea2_example_yoga.png:   3%|3         | 87.2kB / 2.70MB            
+
+
  ...ap_v1.1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
+
  ...swap_v1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
+
+
  ...krea2_example_closeup.png:  19%|#9        |  516kB / 2.66MB            
+
+
+
+
+
  ...krea2_example_kitchen.png:  19%|#9        |  540kB / 2.79MB            
+
+
+
+
+
+
  ...rea2_example_outdoors.png:  19%|#9        |  493kB / 2.54MB            
+
+
+
+
+
+
+
  ...a2_example_restaurant.png:  19%|#9        |  613kB / 3.16MB            
+
+
+
+
+
+
+
+
  ..._krea2_example_seated.png:  19%|#9        |  484kB / 2.50MB            
+
+
+
+
+
+
+
+
+
  ...v1_krea2_example_yoga.png:  19%|#9        |  523kB / 2.70MB            
+
+
+
  ...ap_v1.1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
+
  ...swap_v1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
+
+
  ...krea2_example_closeup.png:  48%|####8     | 1.29MB / 2.66MB            
+
+
+
+
+
  ...krea2_example_kitchen.png:  48%|####8     | 1.35MB / 2.79MB            
+
+
+
+
+
+
  ...rea2_example_outdoors.png:  48%|####8     | 1.23MB / 2.54MB            
+
+
+
+
+
+
+
  ...a2_example_restaurant.png:  48%|####8     | 1.53MB / 3.16MB            
+
+
+
+
+
+
+
+
  ..._krea2_example_seated.png:  48%|####8     | 1.21MB / 2.50MB            
+
+
+
+
+
+
+
+
+
  ...v1_krea2_example_yoga.png:  48%|####8     | 1.31MB / 2.70MB            
+
+
+
  ...ap_v1.1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
+
  ...swap_v1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
+
+
  ...krea2_example_closeup.png:  94%|#########3| 2.50MB / 2.66MB            
+
+
+
+
+
  ...krea2_example_kitchen.png:  94%|#########3| 2.61MB / 2.79MB            
+
+
+
+
+
+
  ...rea2_example_outdoors.png:  94%|#########3| 2.38MB / 2.54MB            
+
+
+
+
+
+
+
  ...a2_example_restaurant.png:  94%|#########3| 2.97MB / 3.16MB            
+
+
+
+
+
+
+
+
  ..._krea2_example_seated.png:  94%|#########3| 2.34MB / 2.50MB            
+
+
+
+
+
+
+
+
+
  ...v1_krea2_example_yoga.png:  94%|#########3| 2.53MB / 2.70MB            
+
+
+
  ...ap_v1.1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
+
  ...swap_v1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
+
+
  ...krea2_example_closeup.png:  97%|#########6| 2.58MB / 2.66MB            
+
+
+
+
+
  ...krea2_example_kitchen.png:  97%|#########6| 2.70MB / 2.79MB            
+
+
+
+
+
+
  ...rea2_example_outdoors.png:  97%|#########6| 2.47MB / 2.54MB            
+
+
+
+
+
+
+
  ...a2_example_restaurant.png:  97%|#########6| 3.07MB / 3.16MB            
+
+
+
+
+
+
+
+
  ..._krea2_example_seated.png:  97%|#########6| 2.42MB / 2.50MB            
+
+
+
+
+
+
+
+
+
  ...v1_krea2_example_yoga.png:  97%|#########6| 2.62MB / 2.70MB            
+
+
+
  ...ap_v1.1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
+
  ...swap_v1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
+
+
  ...krea2_example_closeup.png:  97%|#########6| 2.58MB / 2.66MB            
+
+
+
+
+
  ...krea2_example_kitchen.png:  97%|#########6| 2.70MB / 2.79MB            
+
+
+
+
+
+
  ...rea2_example_outdoors.png:  97%|#########6| 2.47MB / 2.54MB            
+
+
+
+
+
+
+
  ...a2_example_restaurant.png:  97%|#########6| 3.07MB / 3.16MB            
+
+
+
+
+
+
+
+
  ..._krea2_example_seated.png:  97%|#########6| 2.42MB / 2.50MB            
+
+
+
+
+
+
+
+
+
  ...v1_krea2_example_yoga.png:  97%|#########6| 2.62MB / 2.70MB            
+
+
  ...ap_v1.1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
+
  ...swap_v1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
+
+
  ...krea2_example_closeup.png: 100%|##########| 2.66MB / 2.66MB            
+
+
+
+
+
  ...krea2_example_kitchen.png: 100%|##########| 2.79MB / 2.79MB            
+
+
+
+
+
+
  ...rea2_example_outdoors.png: 100%|##########| 2.54MB / 2.54MB            
+
+
+
+
+
+
+
  ...a2_example_restaurant.png: 100%|##########| 3.16MB / 3.16MB            
+
+
+
+
+
+
+
+
  ..._krea2_example_seated.png: 100%|##########| 2.50MB / 2.50MB            
+
+
+
+
+
+
+
+
+
  ...v1_krea2_example_yoga.png: 100%|##########| 2.70MB / 2.70MB            
+
+
+
  ...ap_v1.1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
+
  ...swap_v1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
+
+
  ...krea2_example_closeup.png: 100%|##########| 2.66MB / 2.66MB            
+
+
+
+
+
  ...krea2_example_kitchen.png: 100%|##########| 2.79MB / 2.79MB            
+
+
+
+
+
+
  ...rea2_example_outdoors.png: 100%|##########| 2.54MB / 2.54MB            
+
+
+
+
+
+
+
  ...a2_example_restaurant.png: 100%|##########| 3.16MB / 3.16MB            
+
+
+
+
+
+
+
+
  ..._krea2_example_seated.png: 100%|##########| 2.50MB / 2.50MB            
+
+
+
+
+
+
+
+
+
  ...v1_krea2_example_yoga.png: 100%|##########| 2.70MB / 2.70MB            
+
+
  ...ap_v1.1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
+
  ...swap_v1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
+
+
  ...krea2_example_closeup.png: 100%|##########| 2.66MB / 2.66MB            
+
+
+
+
+
  ...krea2_example_kitchen.png: 100%|##########| 2.79MB / 2.79MB            
+
+
+
+
+
+
  ...rea2_example_outdoors.png: 100%|##########| 2.54MB / 2.54MB            
+
+
+
+
+
+
+
  ...a2_example_restaurant.png: 100%|##########| 3.16MB / 3.16MB            
+
+
+
+
+
+
+
+
  ..._krea2_example_seated.png: 100%|##########| 2.50MB / 2.50MB            
+
+
+
+
+
+
+
+
+
  ...v1_krea2_example_yoga.png: 100%|##########| 2.70MB / 2.70MB            
+
+
  ...ap_v1.1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
+
  ...swap_v1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
+
+
+
  ...krea2_example_closeup.png: 100%|##########| 2.66MB / 2.66MB            
+
+
+
+
+
  ...krea2_example_kitchen.png: 100%|##########| 2.79MB / 2.79MB            
+
+
+
+
+
+
  ...rea2_example_outdoors.png: 100%|##########| 2.54MB / 2.54MB            
+
+
+
+
+
+
+
  ...a2_example_restaurant.png: 100%|##########| 3.16MB / 3.16MB            
+
+
+
+
+
+
+
+
  ..._krea2_example_seated.png: 100%|##########| 2.50MB / 2.50MB            
+
+
+
+
+
+
+
+
+
  ...v1_krea2_example_yoga.png: 100%|##########| 2.70MB / 2.70MB            
+
+
  ...ap_v1.1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
  ...swap_v1_krea2.safetensors: 100%|##########|  914MB /  914MB            
+
  ...krea2_example_closeup.png: 100%|##########| 2.66MB / 2.66MB            
+
  ...krea2_example_kitchen.png: 100%|##########| 2.79MB / 2.79MB            
+
  ...rea2_example_outdoors.png: 100%|##########| 2.54MB / 2.54MB            
+
  ...a2_example_restaurant.png: 100%|##########| 3.16MB / 3.16MB            
+
  ..._krea2_example_seated.png: 100%|##########| 2.50MB / 2.50MB            
+
  ...v1_krea2_example_yoga.png: 100%|##########| 2.70MB / 2.70MB            

hf_upload.out.log

@@ -0,0 +1 @@
+https://huggingface.co/Alissonerdx/BFS-Best-Face-Swap/commit/c758d55502463bc5eb281326c107d0850d7a8b4a