nodes/cubiq-ComfyUI_IPAdapter_plus
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

doesn't assume cuda

Browse Source
This commit is contained in:
matt3o
2023-12-01 10:38:38 +01:00
parent 77554c835d
commit ade7f2005c
2 changed files with 13 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files
IPAdapterPlus.py
+9 -8
View File
@@ -210,13 +210,13 @@ class IPAdapter(nn.Module):
class CrossAttentionPatch:
# forward for patching
def __init__(self, weight, ipadapter, dtype, number, cond, uncond, weight_type, mask=None, sigma_start=0.0, sigma_end=1.0, unfold_batch=False):
def __init__(self, weight, ipadapter, device, dtype, number, cond, uncond, weight_type, mask=None, sigma_start=0.0, sigma_end=1.0, unfold_batch=False):
self.weights = [weight]
self.ipadapters = [ipadapter]
self.conds = [cond]
self.unconds = [uncond]
self.dtype = dtype
self.device = 'cuda'
self.device = device
self.number = number
self.weight_type = [weight_type]
self.masks = [mask]
@@ -227,15 +227,15 @@ class CrossAttentionPatch:
self.k_key = str(self.number*2+1) + "_to_k_ip"
self.v_key = str(self.number*2+1) + "_to_v_ip"
def set_new_condition(self, weight, ipadapter, cond, uncond, dtype, number, weight_type, mask=None, sigma_start=0.0, sigma_end=1.0, unfold_batch=False):
def set_new_condition(self, weight, ipadapter, device, dtype, number, cond, uncond, weight_type, mask=None, sigma_start=0.0, sigma_end=1.0, unfold_batch=False):
self.weights.append(weight)
self.ipadapters.append(ipadapter)
self.conds.append(cond)
self.unconds.append(uncond)
self.masks.append(mask)
self.dtype = dtype
self.device = device
self.weight_type.append(weight_type)
self.device = 'cuda'
self.sigma_start.append(sigma_start)
self.sigma_end.append(sigma_end)
self.unfold_batch.append(unfold_batch)
@@ -245,7 +245,10 @@ class CrossAttentionPatch:
cond_or_uncond = extra_options["cond_or_uncond"]
sigma = extra_options["sigmas"][0].item() if 'sigmas' in extra_options else 999999999.9
with torch.autocast(device_type=self.device, dtype=self.dtype):
# extra options for AnimateDiff
ad_params = extra_options['ad_params'] if "ad_params" in extra_options else None
with torch.autocast(device_type=self.device.type, dtype=self.dtype):
q = n
k = context_attn2
v = value_attn2
@@ -255,9 +258,6 @@ class CrossAttentionPatch:
out = optimized_attention(q, k, v, extra_options["n_heads"])
_, _, lh, lw = extra_options["original_shape"]
# extra options for AnimateDiff
ad_params = extra_options['ad_params'] if "ad_params" in extra_options else None
for weight, cond, uncond, ipadapter, mask, weight_type, sigma_start, sigma_end, unfold_batch in zip(self.weights, self.conds, self.unconds, self.ipadapters, self.masks, self.weight_type, self.sigma_start, self.sigma_end, self.unfold_batch):
if sigma > sigma_start or sigma < sigma_end:
continue
@@ -492,6 +492,7 @@ class IPAdapterApply:
"number": 0,
"weight": self.weight,
"ipadapter": self.ipadapter,
"device": self.device,
"dtype": self.dtype,
"cond": image_prompt_embeds,
"uncond": uncond_image_prompt_embeds,
README.md
+4 -1
View File
@@ -45,6 +45,8 @@ The IPAdapter are very powerful models for image-to-image conditioning. Given a
**:japanese_goblin: [Attention Masking video](https://www.youtube.com/watch?v=vqG1VXKteQg)**
**:movie_camera: [Animation Features video](https://www.youtube.com/watch?v=ddYbhv3WgWw)**
## Installation
Download or git clone this repository inside `ComfyUI/custom_nodes/` directory.
@@ -200,7 +202,8 @@ If you are interested I've also implemented the same features for [Huggingface D
## IPAdapter in the wild
Let me know if you spot the IPAdapter in the wild!
Let me know if you spot the IPAdapter in the wild or tag @latentvision in the video description!
- For German speakers you can find interesting YouTube tutorials on [A Latent Place](https://www.youtube.com/watch?v=rAWn_0YOBU0).
- In Chinese [Introversify](https://www.youtube.com/watch?v=xl8f3oxZgY8)
- [Scott Detweiler](https://www.youtube.com/watch?v=xzGdynQDzsM) covered this extension.