Feat: Add MPS fallback to CUDA support and uv-compatible CLI for FastVLM inference

.gitignore

@@ -26,6 +26,8 @@ share/python-wheels/
 .installed.cfg
 *.egg
 MANIFEST
+checkpoints
+test
 
 # PyInstaller
 #  Usually these files are written by a python script from a template

README.md

@@ -27,7 +27,27 @@ We use LLaVA codebase to train FastVLM variants. In order to train or finetune y
 please follow instructions provided in [LLaVA](https://github.com/haotian-liu/LLaVA) codebase. 
 We provide instructions for running inference with our models.   
 
+
 ### Setup
+#### 🔹 Option 1: Using `uv` (fast Python package manager)
+
+```bash
+# 1. Install uv if not already installed
+pip3 install uv
+
+# 2. Create a virtual environment
+uv venv .venv
+
+# 3. Activate the virtual environment
+source .venv/bin/activate  # On macOS/Linux
+# .venv\\Scripts\\activate  # On Windows (use this instead)
+
+# 4. Sync and install dependencies
+uv sync
+```
+
+#### 🔸 Option 2: Using conda
+
 ```bash
 conda create -n fastvlm python=3.10
 conda activate fastvlm
@@ -56,7 +76,7 @@ bash get_models.sh   # Files will be downloaded to `checkpoints` directory.
 To run inference of PyTorch checkpoint, follow the instruction below
 ```bash
 python predict.py --model-path /path/to/checkpoint-dir \
-                  --image-file /path/to/image.png \
+                  --image-file test/image.png \
                   --prompt "Describe the image."
 ```
 

predict.py

@@ -7,6 +7,7 @@
 
 import torch
 from PIL import Image
+from loguru import logger
 
 from llava.utils import disable_torch_init
 from llava.conversation import conv_templates
@@ -15,9 +16,22 @@
 from llava.constants import IMAGE_TOKEN_INDEX, DEFAULT_IMAGE_TOKEN, DEFAULT_IM_START_TOKEN, DEFAULT_IM_END_TOKEN
 
 
+if torch.backends.mps.is_available():
+    DEVICE = torch.device("mps")
+    logger.info("Using MPS device")
+elif torch.cuda.is_available():
+    DEVICE = torch.device("cuda")
+    logger.info("Using CUDA device")
+else:
+    DEVICE = torch.device("cpu")
+    logger.info("Using CPU device")
+
+
 def predict(args):
-    # Remove generation config from model folder
-    # to read generation parameters from args
+    logger.info(f"Starting prediction with model_path={args.model_path}")
+    logger.info(f"Prompt: {args.prompt}")
+
+    # Remove generation config from model folder to read generation parameters from args
     model_path = os.path.expanduser(args.model_path)
     generation_config = None
     if os.path.exists(os.path.join(model_path, 'generation_config.json')):
@@ -28,7 +42,8 @@ def predict(args):
     # Load model
     disable_torch_init()
     model_name = get_model_name_from_path(model_path)
-    tokenizer, model, image_processor, context_len = load_pretrained_model(model_path, args.model_base, model_name, device="mps")
+    tokenizer, model, image_processor, context_len = load_pretrained_model(model_path, args.model_base, model_name, device=DEVICE)
+    logger.info("Model loaded successfully")
 
     # Construct prompt
     qs = args.prompt
@@ -45,7 +60,7 @@ def predict(args):
     model.generation_config.pad_token_id = tokenizer.pad_token_id
 
     # Tokenize prompt
-    input_ids = tokenizer_image_token(prompt, tokenizer, IMAGE_TOKEN_INDEX, return_tensors='pt').unsqueeze(0).to(torch.device("mps"))
+    input_ids = tokenizer_image_token(prompt, tokenizer, IMAGE_TOKEN_INDEX, return_tensors='pt').unsqueeze(0).to(DEVICE)
 
     # Load and preprocess image
     image = Image.open(args.image_file).convert('RGB')
@@ -65,7 +80,7 @@ def predict(args):
             use_cache=True)
 
         outputs = tokenizer.batch_decode(output_ids, skip_special_tokens=True)[0].strip()
-        print(outputs)
+        logger.success(f"Inference completed. Output: {outputs}")
 
     # Restore generation config
     if generation_config is not None:
@@ -84,4 +99,4 @@ def predict(args):
     parser.add_argument("--num_beams", type=int, default=1)
     args = parser.parse_args()
 
-    predict(args)
+    predict(args)

pyproject.toml

@@ -7,19 +7,34 @@ name = "llava"
 version = "1.2.2.post1"
 description = "Towards GPT-4 like large language and visual assistant."
 readme = "README.md"
-requires-python = ">=3.8"
+requires-python = ">=3.10"
 classifiers = [
     "Programming Language :: Python :: 3",
     "License :: OSI Approved :: Apache Software License",
 ]
 dependencies = [
-    "torch==2.6.0", "torchvision==0.21.0",
-    "transformers==4.48.3", "tokenizers==0.21.0", "sentencepiece==0.1.99", "shortuuid",
-    "accelerate==1.6.0", "peft>=0.10.0,<0.14.0", "bitsandbytes",
-    "pydantic", "markdown2[all]", "numpy==1.26.4", "scikit-learn==1.2.2",
-    "gradio==5.11.0", "requests", "uvicorn", "fastapi",
-    "einops==0.6.1", "einops-exts==0.0.4", "timm==1.0.15",
-    "coremltools==8.2"
+    "torch==2.6.0",
+    "torchvision==0.21.0",
+    "transformers==4.48.3",
+    "tokenizers==0.21.0",
+    "sentencepiece==0.1.99",
+    "shortuuid",
+    "accelerate==1.6.0",
+    "peft>=0.10.0,<0.14.0",
+    "bitsandbytes",
+    "pydantic",
+    "markdown2[all]",
+    "numpy==1.26.4",
+    "scikit-learn==1.2.2",
+    "gradio==5.11.0",
+    "requests",
+    "uvicorn",
+    "fastapi",
+    "einops==0.6.1",
+    "einops-exts==0.0.4",
+    "timm==1.0.15",
+    "coremltools==8.2",
+    "loguru>=0.7.3",
 ]
 
 [project.optional-dependencies]