update chroma transformer approximator init params

Author: DN6

src/diffusers/models/transformers/transformer_chroma.py

@@ -416,7 +416,7 @@ def __init__(
         num_attention_heads: int = 24,
         joint_attention_dim: int = 4096,
         axes_dims_rope: Tuple[int, ...] = (16, 56, 56),
-        approximator_in_factor: int = 16,
+        approximator_num_channels: int = 64,
         approximator_hidden_dim: int = 5120,
         approximator_layers: int = 5,
     ):
@@ -427,11 +427,11 @@ def __init__(
         self.pos_embed = FluxPosEmbed(theta=10000, axes_dim=axes_dims_rope)
 
         self.time_text_embed = ChromaCombinedTimestepTextProjEmbeddings(
-            num_channels=approximator_in_factor,
+            num_channels=approximator_num_channels // 4,
             out_dim=3 * num_single_layers + 2 * 6 * num_layers + 2,
         )
         self.distilled_guidance_layer = ChromaApproximator(
-            in_dim=64,
+            in_dim=approximator_num_channels,
             out_dim=self.inner_dim,
             hidden_dim=approximator_hidden_dim,
             n_layers=approximator_layers,

tests/models/transformers/test_models_transformer_chroma.py

@@ -128,7 +128,7 @@ def prepare_init_args_and_inputs_for_common(self):
             "num_attention_heads": 2,
             "joint_attention_dim": 32,
             "axes_dims_rope": [4, 4, 8],
-            "approximator_in_factor": 32,
+            "approximator_num_channels": 8,
             "approximator_hidden_dim": 16,
             "approximator_layers": 1,
         }