Instructions to use BricksDisplay/phi-2-qnn with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use BricksDisplay/phi-2-qnn with Transformers:

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

pipe = pipeline("text-generation", model="BricksDisplay/phi-2-qnn", trust_remote_code=True)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("BricksDisplay/phi-2-qnn", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("BricksDisplay/phi-2-qnn", trust_remote_code=True)

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use BricksDisplay/phi-2-qnn with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "BricksDisplay/phi-2-qnn"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "BricksDisplay/phi-2-qnn",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/BricksDisplay/phi-2-qnn

SGLang

How to use BricksDisplay/phi-2-qnn 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 "BricksDisplay/phi-2-qnn" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "BricksDisplay/phi-2-qnn",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "BricksDisplay/phi-2-qnn" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "BricksDisplay/phi-2-qnn",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use BricksDisplay/phi-2-qnn with Docker Model Runner:
```
docker model run hf.co/BricksDisplay/phi-2-qnn
```

phi-2-qnn

File size: 869 Bytes

3628cdc

{
  "_name_or_path": "microsoft/phi-2",
  "architectures": [
    "PhiForCausalLM"
  ],
  "attention_dropout": 0.0,
  "auto_map": {
    "AutoConfig": "microsoft/phi-2--configuration_phi.PhiConfig",
    "AutoModelForCausalLM": "microsoft/phi-2--modeling_phi.PhiForCausalLM"
  },
  "bos_token_id": 50256,
  "embd_pdrop": 0.0,
  "eos_token_id": 50256,
  "hidden_act": "gelu_new",
  "hidden_size": 2560,
  "initializer_range": 0.02,
  "intermediate_size": 10240,
  "layer_norm_eps": 1e-05,
  "max_position_embeddings": 2048,
  "model_type": "phi",
  "num_attention_heads": 32,
  "num_hidden_layers": 32,
  "num_key_value_heads": 32,
  "partial_rotary_factor": 0.4,
  "qk_layernorm": false,
  "resid_pdrop": 0.1,
  "rope_scaling": null,
  "rope_theta": 10000.0,
  "tie_word_embeddings": false,
  "transformers_version": "4.39.3",
  "use_cache": true,
  "vocab_size": 51200
}