Merge 2fde495c33 into b08debceca

Signed-off-by: bigcat88 <bigcat88@icloud.com>

.coderabbit.yaml

@@ -4,12 +4,12 @@ early_access: false
 tone_instructions: "Only comment on issues introduced by this PR's changes. Do not flag pre-existing problems in moved, re-indented, or reformatted code."
 
 reviews:
-  profile: "chill"
-  request_changes_workflow: false
+  profile: "assertive"
+  request_changes_workflow: true
   high_level_summary: false
   poem: false
   review_status: false
-  review_details: false
+  review_details: true
   commit_status: true
   collapse_walkthrough: true
   changed_files_summary: false
@@ -39,6 +39,14 @@ reviews:
     - path: "**"
       instructions: |
         IMPORTANT: Only comment on issues directly introduced by this PR's code changes.
+        Treat AGENTS.md as mandatory repository policy, not optional style guidance.
+        Flag PR changes that violate AGENTS.md even when the code is otherwise functional.
+        In particular, enforce architecture boundaries, dtype/device/memory rules,
+        interface contracts, import style, no unnecessary try/except blocks, no inline
+        imports, no outbound internet paths in core ComfyUI, and narrow scoped fixes.
+        Prefer direct findings over suggestions when a rule is violated. Only ignore
+        AGENTS.md when it clearly conflicts with a newer explicit maintainer instruction
+        in the PR.
         Do NOT flag pre-existing issues in code that was merely moved, re-indented,
         de-indented, or reformatted without logic changes. If code appears in the diff
         only due to whitespace or structural reformatting (e.g., removing a `with:` block),
@@ -123,5 +131,10 @@ chat:
 
 knowledge_base:
   opt_out: false
+  code_guidelines:
+    enabled: true
+    filePatterns:
+      - files: "AGENTS.md"
+        applyTo: "**"
   learnings:
     scope: "auto"

comfy/ldm/modules/attention.py

@@ -17,8 +17,9 @@ from comfy import model_management
 TORCH_HAS_GQA = model_management.torch_version_numeric >= (2, 5)
 
 if model_management.xformers_enabled():
-    import xformers
-    import xformers.ops
+    # xFormers's fmha module is now provided by MSLK
+    import mslk
+    import mslk.attention.fmha
 
 SAGE_ATTENTION_IS_AVAILABLE = False
 try:
@@ -415,12 +416,6 @@ def attention_split(q, k, v, heads, mask=None, attn_precision=None, skip_reshape
     return r1
 
 BROKEN_XFORMERS = False
-try:
-    x_vers = xformers.__version__
-    # XFormers bug confirmed on all versions from 0.0.21 to 0.0.26 (q with bs bigger than 65535 gives CUDA error)
-    BROKEN_XFORMERS = x_vers.startswith("0.0.2") and not x_vers.startswith("0.0.20")
-except:
-    pass
 
 @wrap_attn
 def attention_xformers(q, k, v, heads, mask=None, attn_precision=None, skip_reshape=False, skip_output_reshape=False, **kwargs):
@@ -474,7 +469,8 @@ def attention_xformers(q, k, v, heads, mask=None, attn_precision=None, skip_resh
         mask = mask_out[..., :mask.shape[-1]]
         mask = mask.expand(b, heads, -1, -1)
 
-    out = xformers.ops.memory_efficient_attention(q, k, v, attn_bias=mask)
+    # xFormers's fmha module is now provided by MSLK
+    out = mslk.attention.fmha.memory_efficient_attention(q, k, v, attn_bias=mask)
 
     if skip_output_reshape:
         out = out.permute(0, 2, 1, 3)

comfy/ldm/modules/diffusionmodules/model.py

@@ -10,8 +10,8 @@ import comfy.ops
 ops = comfy.ops.disable_weight_init
 
 if model_management.xformers_enabled_vae():
-    import xformers
-    import xformers.ops
+    # xFormers's fmha module is now provided by MSLK
+    import mslk.attention.fmha
 
 def torch_cat_if_needed(xl, dim):
     xl = [x for x in xl if x is not None and x.shape[dim] > 0]
@@ -295,7 +295,8 @@ def xformers_attention(q, k, v):
     )
 
     try:
-        out = xformers.ops.memory_efficient_attention(q, k, v, attn_bias=None)
+        # xFormers's fmha module is now provided by MSLK
+        out = mslk.attention.fmha.memory_efficient_attention(q, k, v, attn_bias=None)
         out = out.transpose(1, 2).reshape(orig_shape)
     except NotImplementedError:
         out = slice_attention(q.view(B, -1, C), k.view(B, -1, C).transpose(1, 2), v.view(B, -1, C).transpose(1, 2)).reshape(orig_shape)

comfy/ldm/pixart/blocks.py

@@ -10,11 +10,9 @@ from comfy.ldm.modules.diffusionmodules.mmdit import TimestepEmbedder, Mlp, time
 from comfy.ldm.modules.attention import optimized_attention
 
 # if model_management.xformers_enabled():
-#     import xformers.ops
-#     if int((xformers.__version__).split(".")[2].split("+")[0]) >= 28:
-#         block_diagonal_mask_from_seqlens = xformers.ops.fmha.attn_bias.BlockDiagonalMask.from_seqlens
-#     else:
-#         block_diagonal_mask_from_seqlens = xformers.ops.fmha.BlockDiagonalMask.from_seqlens
+#     # xFormers's fmha module is now provided by MSLK
+#     import mslk.attention.fmha
+#     block_diagonal_mask_from_seqlens = mslk.attention.fmha.attn_bias.BlockDiagonalMask.from_seqlens
 
 def modulate(x, shift, scale):
     return x * (1 + scale.unsqueeze(1)) + shift.unsqueeze(1)
@@ -51,7 +49,8 @@ class MultiHeadCrossAttention(nn.Module):
         #     attn_bias = None
         #     if mask is not None:
         #         attn_bias = block_diagonal_mask_from_seqlens([N] * B, mask)
-        #     x = xformers.ops.memory_efficient_attention(q, k, v, p=0, attn_bias=attn_bias)
+        #     # xFormers's fmha module is now provided by MSLK
+        #     x = mslk.attention.fmha.memory_efficient_attention(q, k, v, p=0, attn_bias=attn_bias)
         # else:
         #     q, k, v = map(lambda t: t.transpose(1, 2), (q, k, v),)
         #     attn_mask = None

comfy/model_management.py

@@ -399,20 +399,13 @@ if args.disable_xformers:
     XFORMERS_IS_AVAILABLE = False
 else:
     try:
-        import xformers
-        import xformers.ops
+        # xFormers's fmha module is now provided by MSLK
+        import mslk
+        import mslk.attention.fmha
         XFORMERS_IS_AVAILABLE = True
         try:
-            XFORMERS_IS_AVAILABLE = xformers._has_cpp_library
-        except:
-            pass
-        try:
-            XFORMERS_VERSION = xformers.version.__version__
+            XFORMERS_VERSION = mslk.__version__
             logging.info("xformers version: {}".format(XFORMERS_VERSION))
-            if XFORMERS_VERSION.startswith("0.0.18"):
-                logging.warning("\nWARNING: This version of xformers has a major bug where you will get black images when generating high resolution images.")
-                logging.warning("Please downgrade or upgrade xformers to a different version.\n")
-                XFORMERS_ENABLED_VAE = False
         except:
             pass
     except:

comfy/sd1_clip.py

@@ -543,18 +543,24 @@ class SDTokenizer:
     def _try_get_embedding(self, embedding_name:str):
         '''
         Takes a potential embedding name and tries to retrieve it.
-        Returns a Tuple consisting of the embedding and any leftover string, embedding can be None.
+        Returns a Tuple consisting of the embedding, the cleaned embedding name, and any leftover string, embedding can be None.
         '''
         split_embed = embedding_name.split()
         embedding_name = split_embed[0]
         leftover = ' '.join(split_embed[1:])
+
+        match = re.search(r'[<\[]', embedding_name)
+        if match is not None:
+            leftover = embedding_name[match.start():] + (" " + leftover if leftover else "")
+            embedding_name = embedding_name[:match.start()]
+
         embed = load_embed(embedding_name, self.embedding_directory, self.embedding_size, self.embedding_key)
         if embed is None:
             stripped = embedding_name.strip(',')
             if len(stripped) < len(embedding_name):
                 embed = load_embed(stripped, self.embedding_directory, self.embedding_size, self.embedding_key)
-                return (embed, "{} {}".format(embedding_name[len(stripped):], leftover))
-        return (embed, leftover)
+                return (embed, embedding_name, "{} {}".format(embedding_name[len(stripped):], leftover))
+        return (embed, embedding_name, leftover)
 
     def pad_tokens(self, tokens, amount):
         if self.pad_left:
@@ -585,7 +591,7 @@ class SDTokenizer:
         tokens = []
         for weighted_segment, weight in parsed_weights:
             to_tokenize = unescape_important(weighted_segment)
-            split = re.split(' {0}|\n{0}'.format(self.embedding_identifier), to_tokenize)
+            split = re.split(r'(?<=\s){}'.format(re.escape(self.embedding_identifier)), to_tokenize)
             to_tokenize = [split[0]]
             for i in range(1, len(split)):
                 to_tokenize.append("{}{}".format(self.embedding_identifier, split[i]))
@@ -595,7 +601,7 @@ class SDTokenizer:
                 # if we find an embedding, deal with the embedding
                 if word.startswith(self.embedding_identifier) and self.embedding_directory is not None:
                     embedding_name = word[len(self.embedding_identifier):].strip('\n')
-                    embed, leftover = self._try_get_embedding(embedding_name)
+                    embed, embedding_name, leftover = self._try_get_embedding(embedding_name)
                     if embed is None:
                         logging.warning(f"warning, embedding:{embedding_name} does not exist, ignoring")
                     else:

comfy/text_encoders/llama.py

@@ -937,22 +937,41 @@ class BaseGenerate:
             return torch.argmax(logits, dim=-1, keepdim=True)
 
         # Sampling mode
-        if repetition_penalty != 1.0:
-            for i in range(logits.shape[0]):
-                for token_id in set(token_history):
-                    logits[i, token_id] *= repetition_penalty if logits[i, token_id] < 0 else 1/repetition_penalty
-
-        if presence_penalty is not None and presence_penalty != 0.0:
-            for i in range(logits.shape[0]):
-                for token_id in set(token_history):
-                    logits[i, token_id] -= presence_penalty
+        if len(token_history) > 0 and (repetition_penalty != 1.0 or (presence_penalty is not None and presence_penalty != 0.0)):
+            token_ids = torch.tensor(list(set(token_history)), device=logits.device)
+            token_logits = logits[:, token_ids]
+            if repetition_penalty != 1.0:
+                token_logits = torch.where(token_logits < 0, token_logits * repetition_penalty, token_logits / repetition_penalty)
+            if presence_penalty is not None and presence_penalty != 0.0:
+                token_logits = token_logits - presence_penalty
+            logits[:, token_ids] = token_logits
 
         if temperature != 1.0:
             logits = logits / temperature
 
         if top_k > 0:
-            indices_to_remove = logits < torch.topk(logits, top_k)[0][..., -1, None]
-            logits[indices_to_remove] = torch.finfo(logits.dtype).min
+            top_k = min(top_k, logits.shape[-1])
+            logits, top_indices = torch.topk(logits, top_k)
+
+            if min_p > 0.0:
+                probs_before_filter = torch.nn.functional.softmax(logits, dim=-1)
+                top_probs, _ = probs_before_filter.max(dim=-1, keepdim=True)
+                min_threshold = min_p * top_probs
+                indices_to_remove = probs_before_filter < min_threshold
+                logits[indices_to_remove] = torch.finfo(logits.dtype).min
+
+            if top_p < 1.0:
+                sorted_logits, sorted_indices = torch.sort(logits, descending=True)
+                cumulative_probs = torch.cumsum(torch.nn.functional.softmax(sorted_logits, dim=-1), dim=-1)
+                sorted_indices_to_remove = cumulative_probs > top_p
+                sorted_indices_to_remove[..., 0] = False
+                indices_to_remove = torch.zeros_like(logits, dtype=torch.bool)
+                indices_to_remove.scatter_(1, sorted_indices, sorted_indices_to_remove)
+                logits[indices_to_remove] = torch.finfo(logits.dtype).min
+
+            probs = torch.nn.functional.softmax(logits, dim=-1)
+            next_token = torch.multinomial(probs, num_samples=1, generator=generator)
+            return top_indices.gather(1, next_token)
 
         if min_p > 0.0:
             probs_before_filter = torch.nn.functional.softmax(logits, dim=-1)

comfy_api_nodes/util/request_logger.py

@@ -9,6 +9,7 @@ from typing import Any
 import folder_paths
 
 logger = logging.getLogger(__name__)
+_SENSITIVE_HEADERS = {"authorization", "x-api-key"}
 
 
 def get_log_directory():
@@ -73,6 +74,10 @@ def _format_data_for_logging(data: Any) -> str:
     return str(data)
 
 
+def _redact_headers(headers: dict) -> dict:
+    return {k: ("***" if k.lower() in _SENSITIVE_HEADERS else v) for k, v in headers.items()}
+
+
 def log_request_response(
     operation_id: str,
     request_method: str,
@@ -101,7 +106,7 @@ def log_request_response(
         log_content.append(f"Method: {request_method}")
         log_content.append(f"URL: {request_url}")
         if request_headers:
-            log_content.append(f"Headers:\n{_format_data_for_logging(request_headers)}")
+            log_content.append(f"Headers:\n{_format_data_for_logging(_redact_headers(request_headers))}")
         if request_params:
             log_content.append(f"Params:\n{_format_data_for_logging(request_params)}")
         if request_data is not None:

comfy_extras/nodes_color.py

@@ -16,23 +16,30 @@ class ColorToRGBInt(io.ComfyNode):
             ],
             outputs=[
                 io.Int.Output(display_name="rgb_int"),
-                io.Color.Output(display_name="hex")
+                io.Color.Output(display_name="hex"),
+                io.Float.Output(display_name="alpha"),
             ],
         )
 
     @classmethod
     def execute(cls, color: str) -> io.NodeOutput:
-        # expect format #RRGGBB
-        if len(color) != 7 or color[0] != "#":
-            raise ValueError("Color must be in format #RRGGBB")
+        # expect format #RRGGBB or #RRGGBBAA
+        if len(color) not in (7, 9) or color[0] != "#":
+            raise ValueError("Color must be in format #RRGGBB or #RRGGBBAA")
         try:
             int(color[1:], 16)
         except ValueError:
-            raise ValueError("Color must be in format #RRGGBB") from None
+            raise ValueError("Color must be in format #RRGGBB or #RRGGBBAA") from None
+
+        alpha = 1.0
+        if len(color) == 9:
+            alpha = int(color[7:9], 16) / 255.0
+            color = color[:7]
+
         r, g, b = hex_to_rgb(color)
 
         rgb_int = r * 256 * 256 + g * 256 + b
-        return io.NodeOutput(rgb_int, color)
+        return io.NodeOutput(rgb_int, color, alpha)
 
 
 class ColorExtension(ComfyExtension):

requirements.txt

@@ -1,6 +1,6 @@
 comfyui-frontend-package==1.45.20
 comfyui-workflow-templates==0.11.2
-comfyui-embedded-docs==0.5.6
+comfyui-embedded-docs==0.5.7
 torch
 torchsde
 torchvision