init

configs/models/qwen2_5_1_5b_radio_sd3_dynamic_puffin copy.py

@@ -0,0 +1,90 @@
+import torch
+from src.models.puffin.model import Qwen2p5RadioStableDiffusion3HFDynamic
+from src.models.stable_diffusion3.transformer_sd3_dynamic import SD3Transformer2DModel
+from src.models.radiov3.hf_model import RADIOModel
+from diffusers import AutoencoderKL, FlowMatchEulerDiscreteScheduler
+from transformers import AutoModelForCausalLM, AutoTokenizer
+
+llm_name_or_path = 'Qwen/Qwen2.5-1.5B-Instruct'
+sd3_model_name_or_path = "stabilityai/stable-diffusion-3-medium-diffusers"
+
+prompt_template = dict(
+    SYSTEM=('<|im_start|>system\n{system}<|im_end|>\n'),
+    INSTRUCTION=('<|im_start|>user\n{input}<|im_end|>\n'
+                 '<|im_start|>assistant\n'),
+    SUFFIX='<|im_end|>',
+    IMG_START_TOKEN='<|vision_start|>',
+    IMG_END_TOKEN='<|vision_end|>',
+    IMG_CONTEXT_TOKEN='<|image_pad|>',
+    GENERATION='Generate an image: {input}',
+    GENERATION_CROSS='Generate a target image given an initial view: {input}',
+    SUFFIX_AS_EOS=True,
+    SEP='\n',
+    STOP_WORDS=['<|im_end|>', '<|endoftext|>']
+)
+
+model = dict(type=Qwen2p5RadioStableDiffusion3HFDynamic,
+             num_queries=64,
+             connector_1=dict(
+                 hidden_size=1024,
+                 intermediate_size=4096,
+                 num_hidden_layers=6,
+                 #_attn_implementation='flash_attention_2',
+                 num_attention_heads=16, ),
+             connector_2=dict(
+                 hidden_size=1024,
+                 intermediate_size=4096,
+                 num_hidden_layers=6,
+                 #_attn_implementation='flash_attention_2',
+                 num_attention_heads=16, 
+                 ),
+             transformer=dict(
+                 type=SD3Transformer2DModel.from_pretrained,
+                 pretrained_model_name_or_path=sd3_model_name_or_path,
+                 subfolder="transformer",
+                 torch_dtype=torch.bfloat16,
+                 #local_files_only=True,
+                 ),
+             test_scheduler=dict(
+                 type=FlowMatchEulerDiscreteScheduler.from_pretrained,
+                 pretrained_model_name_or_path=sd3_model_name_or_path,
+                 subfolder="scheduler",
+                 #local_files_only=True,
+                 ),
+             train_scheduler=dict(
+                 type=FlowMatchEulerDiscreteScheduler.from_pretrained,
+                 pretrained_model_name_or_path=sd3_model_name_or_path,
+                 subfolder="scheduler",
+                 #local_files_only=True,
+                 ),
+             vae=dict(
+                 type=AutoencoderKL.from_pretrained,
+                 pretrained_model_name_or_path=sd3_model_name_or_path,
+                 subfolder="vae",
+                 torch_dtype=torch.bfloat16,
+                 #local_files_only=True,
+                 ),
+             freeze_visual_encoder=True,
+             freeze_llm=True,
+             llm=dict(
+                 type=AutoModelForCausalLM.from_pretrained,
+                 pretrained_model_name_or_path=llm_name_or_path,
+                 torch_dtype=torch.bfloat16,
+                 #local_files_only=True,
+                 #attn_implementation='flash_attention_2',
+             ),
+             tokenizer=dict(
+                 type=AutoTokenizer.from_pretrained,
+                 pretrained_model_name_or_path=llm_name_or_path,
+                 #local_files_only=True,
+                 ),
+             prompt_template=prompt_template,
+             pretrained_pth=None,
+             use_activation_checkpointing=False,
+             visual_encoder=dict(
+                 type=RADIOModel.from_pretrained,
+                 pretrained_model_name_or_path="nvidia/C-RADIOv3-H",
+                 torch_dtype=torch.bfloat16,
+                 #local_files_only=True,
+                 ),
+             )

configs/models/qwen2_5_1_5b_radio_sd3_dynamic_puffin.py

@@ -38,13 +38,7 @@ model = dict(type=Qwen2p5RadioStableDiffusion3HFDynamic,
                  #_attn_implementation='flash_attention_2',
                  num_attention_heads=16, 
                  ),
-             transformer=dict(
-                 type=SD3Transformer2DModel.from_pretrained,
-                 pretrained_model_name_or_path=sd3_model_name_or_path,
-                 subfolder="transformer",
-                 torch_dtype=torch.bfloat16,
-                 #local_files_only=True,
-                 ),
+             transformer=SD3Transformer2DModel,
              test_scheduler=dict(
                  type=FlowMatchEulerDiscreteScheduler.from_pretrained,
                  pretrained_model_name_or_path=sd3_model_name_or_path,