diff --git a/api/onnx_web/convert/__main__.py b/api/onnx_web/convert/__main__.py
index 8bd5d1a1..b7515b98 100644
--- a/api/onnx_web/convert/__main__.py
+++ b/api/onnx_web/convert/__main__.py
@@ -269,7 +269,12 @@ def convert_model_diffusion(conversion: ConversionContext, model):
     model_format = source_format(model)
 
     pipeline = model.get("pipeline", "txt2img")
+    logger.trace("converting diffusion model using pipeline %s", pipeline)
+
     converter = model_converters.get(pipeline)
+    if converter is None:
+        raise ValueError("cannot find converter for pipeline")
+
     converted, dest = converter(
         conversion,
         model,
@@ -500,7 +505,7 @@ def register_plugins(conversion: ConversionContext):
     logger.info("loading conversion plugins")
     exports = load_plugins(conversion)
 
-    for proto, client in exports.clients:
+    for proto, client in exports.clients.items():
         try:
             add_model_source(proto, client)
         except Exception:
diff --git a/api/onnx_web/diffusers/utils.py b/api/onnx_web/diffusers/utils.py
index 1c6f8ed0..d83d0f98 100644
--- a/api/onnx_web/diffusers/utils.py
+++ b/api/onnx_web/diffusers/utils.py
@@ -198,27 +198,29 @@ def expand_prompt(
         negative_prompt_embeds = self.text_encoder(
             input_ids=uncond_input.input_ids.astype(np.int32)
         )[0]
-        negative_padding = tokens.input_ids.shape[1] - negative_prompt_embeds.shape[1]
-        logger.trace(
-            "padding negative prompt to match input: %s, %s, %s extra tokens",
-            tokens.input_ids.shape,
-            negative_prompt_embeds.shape,
-            negative_padding,
-        )
-        negative_prompt_embeds = np.pad(
-            negative_prompt_embeds,
-            [(0, 0), (0, negative_padding), (0, 0)],
-            mode="constant",
-            constant_values=0,
-        )
-        negative_prompt_embeds = np.repeat(
-            negative_prompt_embeds, num_images_per_prompt, axis=0
-        )
 
-        # For classifier free guidance, we need to do two forward passes.
-        # Here we concatenate the unconditional and text embeddings into a single batch
-        # to avoid doing two forward passes
-        prompt_embeds = np.concatenate([negative_prompt_embeds, prompt_embeds])
+        if negative_prompt_embeds is not None:
+            negative_padding = tokens.input_ids.shape[1] - negative_prompt_embeds.shape[1]
+            logger.trace(
+                "padding negative prompt to match input: %s, %s, %s extra tokens",
+                tokens.input_ids.shape,
+                negative_prompt_embeds.shape,
+                negative_padding,
+            )
+            negative_prompt_embeds = np.pad(
+                negative_prompt_embeds,
+                [(0, 0), (0, negative_padding), (0, 0)],
+                mode="constant",
+                constant_values=0,
+            )
+            negative_prompt_embeds = np.repeat(
+                negative_prompt_embeds, num_images_per_prompt, axis=0
+            )
+
+            # For classifier free guidance, we need to do two forward passes.
+            # Here we concatenate the unconditional and text embeddings into a single batch
+            # to avoid doing two forward passes
+            prompt_embeds = np.concatenate([negative_prompt_embeds, prompt_embeds])
 
     logger.trace("expanded prompt shape: %s", prompt_embeds.shape)
     return prompt_embeds
@@ -427,7 +429,7 @@ def parse_wildcards(prompt: str, seed: int, wildcards: Dict[str, List[str]]) ->
 
         wildcard = ""
         if name in wildcards:
-            wildcard = pop_random(wildcards.get(name))
+            wildcard = pop_random(wildcards[name])
         else:
             logger.warning("unknown wildcard: %s", name)
 
@@ -488,11 +490,11 @@ def parse_region_group(group: Tuple[str, ...]) -> Region:
     top, left, bottom, right, weight, feather, prompt = group
 
     # break down the feather section
-    feather_radius, *feather_edges = feather.split("_")
-    if len(feather_edges) == 0:
+    feather_radius, *feather_rest = feather.split("_")
+    if len(feather_rest) == 0:
         feather_edges = "TLBR"
     else:
-        feather_edges = "".join(feather_edges)
+        feather_edges = "".join(feather_rest)
 
     return (
         int(top),
diff --git a/api/onnx_web/worker/pool.py b/api/onnx_web/worker/pool.py
index 3b0d32a8..d210421e 100644
--- a/api/onnx_web/worker/pool.py
+++ b/api/onnx_web/worker/pool.py
@@ -536,7 +536,7 @@ def logger_main(pool: DevicePoolExecutor, logs: "Queue[str]"):
 
     while True:
         try:
-            msg = logs.get(pool.join_timeout / 2)
+            msg = logs.get(timeout=(pool.join_timeout / 2))
             logger.debug("received logs from worker: %s", msg)
         except Empty:
             # logger worker should not generate more logs if it doesn't have any logs