fix prompt caching

Signed-off-by: Vladimir Mandic <mandic00@live.com>
This commit is contained in:
Vladimir Mandic
2026-09-04 15:47:06 +02:00
parent 653fbf4d51
commit 0dbe27c371
2 changed files with 22 additions and 8 deletions
+1
View File
@@ -73,6 +73,7 @@ All-about-optimizations:
- openvino: optimize recompile checks and lora loading
- log: ansi color handling
- compile: keep model compiled state
- prompt: cache checks when cfg changes
## Update for 2026-08-26
+21 -8
View File
@@ -11,29 +11,39 @@ class PromptCache:
self.id = None
self.max = 16
def get(self, prompt):
@staticmethod
def _hashable(val):
if isinstance(val, list):
return tuple(val)
return val
def get(self, prompt, negative_prompt=None, cfg_enabled=None):
if self.id != id(shared.sd_model):
self.cache.clear()
self.id = id(shared.sd_model)
log.debug(f'Encode: prompt cache activate id={self.id} depth={len(self.cache)}')
prompt = self._hashable(prompt)
negative_prompt = self._hashable(negative_prompt)
if (isinstance(prompt, list) and len(prompt) == 1 and isinstance(prompt[0], str)):
cached = self.cache.get(prompt[0], None)
cached = self.cache.get((prompt[0], negative_prompt, cfg_enabled), None)
elif isinstance(prompt, str):
cached = self.cache.get(prompt, None)
cached = self.cache.get((prompt, negative_prompt, cfg_enabled), None)
else:
cached = None
if cached:
log.debug(f'Encode: prompt="{prompt}" cache={len(self.cache)} hit')
return cached
def set(self, prompt, encoded):
def set(self, prompt, encoded, negative_prompt=None, cfg_enabled=None):
if len(self.cache) >= self.max:
oldest_key = next(iter(self.cache))
del self.cache[oldest_key]
prompt = self._hashable(prompt)
negative_prompt = self._hashable(negative_prompt)
if (isinstance(prompt, list) and len(prompt) == 1 and isinstance(prompt[0], str)):
self.cache[prompt[0]] = encoded
self.cache[(prompt[0], negative_prompt, cfg_enabled)] = encoded
elif isinstance(prompt, str):
self.cache[prompt] = encoded
self.cache[(prompt, negative_prompt, cfg_enabled)] = encoded
prompt_cache = PromptCache()
@@ -61,7 +71,10 @@ def hijack_encode_prompt(*args, **kwargs):
if patch_prompt:
args_copy[0] = res
cached = prompt_cache.get(prompt)
# cache key must include cfg-affecting kwargs since encode_prompt output (e.g. negative_prompt_embeds) depends on them
negative_prompt = kwargs.get('negative_prompt', None)
cfg_enabled = kwargs.get('do_classifier_free_guidance', None)
cached = prompt_cache.get(prompt, negative_prompt, cfg_enabled)
if cached is not None:
res = cached
else:
@@ -71,7 +84,7 @@ def hijack_encode_prompt(*args, **kwargs):
res = shared.sd_model.orig_encode_prompt(*args_copy, **kwargs)
else:
res = shared.sd_model.encode_prompt(*args_copy, **kwargs)
prompt_cache.set(prompt, res)
prompt_cache.set(prompt, res, negative_prompt, cfg_enabled)
if hasattr(shared.sd_model, 'after_prompt_encode'):
log.debug(f'Encode: prompt="{prompt}" op=after')