Upload folder using huggingface_hub

animatediff/utils/convert_from_ckpt.py

@@ -160,6 +160,10 @@ def assign_to_checkpoint(
     """
     assert isinstance(paths, list), "Paths should be a list of dicts containing 'old' and 'new' keys."
 
+    
+    
+
+
     # Splits the attention layers into three variables.
     if attention_paths_to_split is not None:
         for path, path_map in attention_paths_to_split.items():
@@ -625,7 +629,10 @@ def convert_ldm_vae_checkpoint(checkpoint, config):
     mid_attentions = [key for key in vae_state_dict if "encoder.mid.attn" in key]
     paths = renew_vae_attention_paths(mid_attentions)
     meta_path = {"old": "mid.attn_1", "new": "mid_block.attentions.0"}
-    assign_to_checkpoint(paths, new_checkpoint, vae_state_dict, additional_replacements=[meta_path], config=config)
+    oldKey = {"old": "key", "new": "to_k"}
+    oldQuery = {"old": "query", "new": "to_q"}
+    oldValue = {"old": "value", "new": "to_v"}
+    assign_to_checkpoint(paths, new_checkpoint, vae_state_dict, additional_replacements=[meta_path, oldKey, oldQuery, oldValue], config=config)
     conv_attn_to_linear(new_checkpoint)
 
     for i in range(num_up_blocks):
@@ -658,7 +665,10 @@ def convert_ldm_vae_checkpoint(checkpoint, config):
     mid_attentions = [key for key in vae_state_dict if "decoder.mid.attn" in key]
     paths = renew_vae_attention_paths(mid_attentions)
     meta_path = {"old": "mid.attn_1", "new": "mid_block.attentions.0"}
-    assign_to_checkpoint(paths, new_checkpoint, vae_state_dict, additional_replacements=[meta_path], config=config)
+    oldKey = {"old": "key", "new": "to_k"}
+    oldQuery = {"old": "query", "new": "to_q"}
+    oldValue = {"old": "value", "new": "to_v"}
+    assign_to_checkpoint(paths, new_checkpoint, vae_state_dict, additional_replacements=[meta_path, oldKey, oldQuery, oldValue], config=config)
     conv_attn_to_linear(new_checkpoint)
     return new_checkpoint
 

data/data.ipynb

@@ -2436,6 +2436,35 @@
    "source": [
     "!ffmpeg -i output/0.mp4"
    ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "config_yaml = f\"\"\"\n",
+    "City:\n",
+    "  base: \"\"\n",
+    "  path: \"\"\n",
+    "  motion_module:\n",
+    "    - \"/mm.pth\"\n",
+    "\n",
+    "\n",
+    "  seed:           [1]\n",
+    "  steps:          50\n",
+    "  guidance_scale: 16\n",
+    "\n",
+    "  prompt:\n",
+    "    - f\"{test_prompt}, masterpiece, best quality, empty New York City street, tall buildings, no people, high definition\"\n",
+    "\n",
+    "  n_prompt:\n",
+    "    - \"\"\n",
+    "\"\"\"\n",
+    "\n",
+    "with open(f'configs/prompts/.yaml', 'w') as f:\n",
+    "  f.write(config_yaml)"
+   ]
   }
  ],
  "metadata": {