[Model] Add circlestone-labs/Anima (#4083)

Signed-off-by: akshatvishu <akshatnayak197@gmail.com>

examples/offline_inference/text_to_image/README.md

@@ -19,23 +19,23 @@ This folder provides several entrypoints for experimenting with text-to-image di
 
 ### Supported Models
 
-| Model | Image Shape  | Peak VRAM (GiB) * | Model Weights (GiB) |
+| Model | Image Shape | Peak VRAM (GiB) * | Model Weights (GiB) |
 | ----- | ----------- | ----------- | ----------------- |
 | `Qwen/Qwen-Image` | 1024 x 1024 | 60.0 | 53.7 |
-| `Qwen/Qwen-Image-2512` |1024 x 1024 | 60.0 | 53.7 |
+| `Qwen/Qwen-Image-2512` | 1024 x 1024 | 60.0 | 53.7 |
 | `Tongyi-MAI/Z-Image-Turbo` | 1024 x 1024 | 24.8 | 19.2 |
 | `stepfun-ai/NextStep-1.1` | 512 x 512 | 71.8 | 28.1 |
 | `meituan-longcat/LongCat-Image` | 1024 x 1024 | 71.2 | 27.3 |
 | `AIDC-AI/Ovis-Image-7B` | 1024 x 1024 | 71.8 | 17.1 |
-| `OmniGen2/OmniGen2` |  1024 x 1024 | 20.1 | 14.7 |
+| `OmniGen2/OmniGen2` | 1024 x 1024 | 20.1 | 14.7 |
 | `stabilityai/stable-diffusion-3.5-medium` | 1024 x 1024 | 20.1 | 15.6 |
 | `black-forest-labs/FLUX.1-dev` | 1024 x 1024 | 33.9 | 31.4 |
 | `black-forest-labs/FLUX.1-schnell` | 1024 x 1024 | 33.9 | 31.4 |
 | `black-forest-labs/FLUX.2-klein-4B` | 1024 x 1024 | 72.7 | 14.9 |
 | `black-forest-labs/FLUX.2-klein-9B` | 1024 x 1024 | 37.1 | 32.3 |
 | `black-forest-labs/FLUX.2-dev` | 1024 x 1024 | 65.7 | >80 (CPU offload required) |
-| `HunyuanImage-3.0` | 1024 x 1024 | 80.0 (TP≥3)  | 160 |
-| `HiDream-I1-Full` | 1024 x 1024 | 63.7  | 57.7 |
+| `HunyuanImage-3.0` | 1024 x 1024 | 80.0 (TP≥3) | 160 |
+| `HiDream-I1-Full` | 1024 x 1024 | 63.7 | 57.7 |
 | `krea/Krea-2-Raw`, `krea/Krea-2-Turbo` | 1024 (Raw) / 2048 (Turbo) | — | ~30 |
 
 !!! info
@@ -112,9 +112,9 @@ python text_to_image.py \
 | `--enable-cpu-offload` | flag | off | Enable CPU offloading for diffusion models |
 | `--lora-path` | str | — | Path to PEFT LoRA adapter folder or checkpoint file |
 | `--lora-scale` | float | `1.0` | Scale factor for LoRA weights |
-| `--lora-backend` | str |`"peft"`| LoRA backend for loading LoRA adapters. Default: peft. Available options: peft, distill |
-| `--use-system-prompt` | str | `None` | System prompt preset: `en_unified`, `en_vanilla`, `en_recaption`, `en_think_recaption`, `dynamic`, `None`, or custom text. Recommended: `en_unified`. Only for HunyuanImage-3.0.|
-| `--system-prompt` | str | `None` | Custom system prompt text. Only used when `--use-system-prompt` is set to `custom`. Only for HunyuanImage-3.0.|
+| `--lora-backend` | str | `"peft"` | LoRA backend for loading LoRA adapters. Default: peft. Available options: peft, distill |
+| `--use-system-prompt` | str | `None` | System prompt preset: `en_unified`, `en_vanilla`, `en_recaption`, `en_think_recaption`, `dynamic`, `None`, or custom text. Recommended: `en_unified`. Only for HunyuanImage-3.0. |
+| `--system-prompt` | str | `None` | Custom system prompt text. Only used when `--use-system-prompt` is set to `custom`. Only for HunyuanImage-3.0. |
 | `--auxiliary-text-encoder` | str | `None` | Supplementary auxiliary text encoder parameters model name or path (especially for Hidream-l1-full). |
 
 **NextStep-1.1 specific arguments:**
@@ -127,7 +127,7 @@ python text_to_image.py \
 | `--use-norm` | flag | off | Apply layer normalization to sampled tokens |
 
 > If you encounter OOM errors, try using `--vae-use-slicing` and `--vae-use-tiling` to reduce memory usage.
-
+>
 > Qwen-Image currently publishes best-effort presets at `1328x1328`, `1664x928`, `928x1664`, `1472x1140`, `1140x1472`, `1584x1056`, and `1056x1584`. Adjust `--height/--width` accordingly for the most reliable outcomes.
 
 ## More CLI Examples
@@ -216,6 +216,26 @@ python examples/offline_inference/text_to_image/text_to_image.py \
   --output /output.png
 ```
 
+### Anima Single-File Checkpoints
+
+Anima uses the official model dimensions by default. A checkpoint may provide a JSON file with the same stem (for example, `anima.json` beside `anima.safetensors`) containing `transformer` and `text_conditioner` configuration objects. These override the default dimensions and must match the checkpoint weights. The shared tiny-model tests use this to load small random-weight models. Components in `text_encoder/`, `vae/`, `tokenizer/`, `t5_tokenizer/`, and `scheduler/` beside the checkpoint are discovered automatically.
+
+To load Anima, point `--model` to the single-file checkpoint path, pass the native pipeline class name using `--model-class-name`, and supply the converted components directory using `--custom-pipeline-args`:
+
+```bash
+python examples/offline_inference/text_to_image/text_to_image.py \
+  --model /path/to/models/anima-official/split_files/diffusion_models/anima-base-v1.0.safetensors \
+  --model-class-name AnimaPipeline \
+  --custom-pipeline-args '{"components_path": "/path/to/models/anima-components"}' \
+  --prompt "A cinematic close-up of a glass teapot on a wooden table." \
+  --seed 42 \
+  --guidance-scale 4.0 \
+  --num-inference-steps 50 \
+  --height 1024 \
+  --width 1024 \
+  --output anima_output.png
+```
+
 ### Krea 2
 
 Krea 2 is published as diffusers-format checkpoints: `krea/Krea-2-Turbo` (few-step distilled) and `krea/Krea-2-Raw`.
@@ -336,7 +356,7 @@ python text_to_image.py \
 
 LoRA adapters must be in PEFT format. A typical adapter directory structure:
 
-```
+```text
 lora_adapter/
 ├── adapter_config.json
 └── adapter_model.safetensors

examples/offline_inference/text_to_image/text_to_image.py

@@ -1,5 +1,5 @@
 # SPDX-License-Identifier: Apache-2.0
-# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
+# SPDX-FileCopyrightText: Copyright contributors to the vLLM-Omni project
 
 import argparse
 import functools
@@ -58,6 +58,7 @@ def parse_json_object(value: str, flag_name: str = "argument") -> dict[str, Any]
 
 
 parse_profiler_config = functools.partial(parse_json_object, flag_name="--profiler-config")
+parse_custom_pipeline_args = functools.partial(parse_json_object, flag_name="--custom-pipeline-args")
 
 
 def build_text_to_image_prompt(prompt: str, negative_prompt: str | None) -> dict[str, Any]:
@@ -379,6 +380,18 @@ def parse_args() -> argparse.Namespace:
         default=None,
         help="Supplementary auxiliary text encoder parameters model name or path (especially for Hidream-l1-full).",
     )
+    parser.add_argument(
+        "--model-class-name",
+        type=str,
+        default=None,
+        help="Override the diffusion pipeline class name (e.g. AnimaPipeline).",
+    )
+    parser.add_argument(
+        "--custom-pipeline-args",
+        type=parse_custom_pipeline_args,
+        default=None,
+        help='JSON object passed to native/custom pipelines (e.g. \'{"components_path": "/path"}\').',
+    )
     parser.add_argument(
         "--trust-remote-code",
         action="store_true",
@@ -576,9 +589,13 @@ def main():
         omni_kwargs["trust_remote_code"] = True
     if args.deploy_config:
         omni_kwargs["deploy_config"] = args.deploy_config
-    if use_nextstep:
+    if args.model_class_name:
+        omni_kwargs["model_class_name"] = args.model_class_name
+    elif use_nextstep:
         # NextStep-1.1 requires explicit pipeline class
         omni_kwargs["model_class_name"] = "NextStep11Pipeline"
+    if args.custom_pipeline_args is not None:
+        omni_kwargs["custom_pipeline_args"] = args.custom_pipeline_args
     # Cosmos3 loads its (gated) guardrail models at build time, so the guardrails
     # gate is an engine-level config (offline analog of the server's --no-guardrails).
     if args.extra_body and "guardrails" in args.extra_body:
@@ -614,6 +631,10 @@ def main():
         print(f"  LoRA: scale={args.lora_scale}")
     if args.deploy_config:
         print(f"  deploy-config: {args.deploy_config}")
+    if args.model_class_name:
+        print(f"  Model class name: {args.model_class_name}")
+    if args.custom_pipeline_args is not None:
+        print(f"  Custom pipeline args: {args.custom_pipeline_args}")
     print(f"{'=' * 60}\n")
 
     # Build LoRA request when --lora-path is set