Instructions to use robot-learning-group47/eval3_lora_fixed_Chengming with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use robot-learning-group47/eval3_lora_fixed_Chengming with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("HuggingFaceTB/SmolVLM2-2.2B-Instruct")
model = PeftModel.from_pretrained(base_model, "robot-learning-group47/eval3_lora_fixed_Chengming")

Transformers

How to use robot-learning-group47/eval3_lora_fixed_Chengming with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="robot-learning-group47/eval3_lora_fixed_Chengming")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("robot-learning-group47/eval3_lora_fixed_Chengming", dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use robot-learning-group47/eval3_lora_fixed_Chengming with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "robot-learning-group47/eval3_lora_fixed_Chengming"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "robot-learning-group47/eval3_lora_fixed_Chengming",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/robot-learning-group47/eval3_lora_fixed_Chengming

SGLang

How to use robot-learning-group47/eval3_lora_fixed_Chengming with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "robot-learning-group47/eval3_lora_fixed_Chengming" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "robot-learning-group47/eval3_lora_fixed_Chengming",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "robot-learning-group47/eval3_lora_fixed_Chengming" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "robot-learning-group47/eval3_lora_fixed_Chengming",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use robot-learning-group47/eval3_lora_fixed_Chengming with Docker Model Runner:
```
docker model run hf.co/robot-learning-group47/eval3_lora_fixed_Chengming
```

eval3_lora_fixed_Chengming

File size: 1,837 Bytes

8e70365

{
  "checkpoint": "/home/shadeform/checkpoint/eval3_phase2_TOY",
  "data_dir": "/home/shadeform/data/eval3_position_qa_3frames",
  "output_dir": "/home/shadeform/outputs/vlm_lora/eval3_phase2_TOY_position_qa_3frames",
  "train_examples": 216,
  "eval_examples": 54,
  "epochs": 5,
  "batch_size": 1,
  "grad_accum": 8,
  "lr": 5e-05,
  "global_step": 135,
  "eval": {
    "examples": 54,
    "correct": 54,
    "accuracy": 1.0,
    "grouped": {
      "target_person": {
        "Barack Obama": {
          "correct": 18,
          "total": 18,
          "accuracy": 1.0,
          "prediction_counts": {
            "left": 6,
            "middle": 6,
            "right": 6
          }
        },
        "Taylor Swift": {
          "correct": 18,
          "total": 18,
          "accuracy": 1.0,
          "prediction_counts": {
            "left": 6,
            "middle": 6,
            "right": 6
          }
        },
        "Yann LeCun": {
          "correct": 18,
          "total": 18,
          "accuracy": 1.0,
          "prediction_counts": {
            "left": 6,
            "middle": 6,
            "right": 6
          }
        }
      },
      "ground_truth": {
        "left": {
          "correct": 18,
          "total": 18,
          "accuracy": 1.0,
          "prediction_counts": {
            "left": 18
          }
        },
        "middle": {
          "correct": 18,
          "total": 18,
          "accuracy": 1.0,
          "prediction_counts": {
            "middle": 18
          }
        },
        "right": {
          "correct": 18,
          "total": 18,
          "accuracy": 1.0,
          "prediction_counts": {
            "right": 18
          }
        }
      }
    },
    "output_csv": "/home/shadeform/outputs/vlm_lora/eval3_phase2_TOY_position_qa_3frames/eval_predictions.csv"
  }
}