Instructions to use jerryzh168/phi4-float8dq with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use jerryzh168/phi4-float8dq with Transformers:

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

pipe = pipeline("text-generation", model="jerryzh168/phi4-float8dq")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("jerryzh168/phi4-float8dq")
model = AutoModelForCausalLM.from_pretrained("jerryzh168/phi4-float8dq", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use jerryzh168/phi4-float8dq with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "jerryzh168/phi4-float8dq"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "jerryzh168/phi4-float8dq",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/jerryzh168/phi4-float8dq

SGLang

How to use jerryzh168/phi4-float8dq 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 "jerryzh168/phi4-float8dq" \
    --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": "jerryzh168/phi4-float8dq",
		"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 "jerryzh168/phi4-float8dq" \
        --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": "jerryzh168/phi4-float8dq",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use jerryzh168/phi4-float8dq with Docker Model Runner:
```
docker model run hf.co/jerryzh168/phi4-float8dq
```

phi4-float8dq

File size: 1,736 Bytes

63e59af

{
  "architectures": [
    "Phi3ForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "bos_token_id": 100257,
  "embd_pdrop": 0.0,
  "eos_token_id": 100265,
  "hidden_act": "silu",
  "hidden_size": 5120,
  "initializer_range": 0.02,
  "intermediate_size": 17920,
  "max_position_embeddings": 16384,
  "model_type": "phi3",
  "num_attention_heads": 40,
  "num_hidden_layers": 40,
  "num_key_value_heads": 10,
  "original_max_position_embeddings": 16384,
  "pad_token_id": 100349,
  "partial_rotary_factor": 1.0,
  "quantization_config": {
    "modules_to_not_convert": null,
    "quant_method": "torchao",
    "quant_type": {
      "default": {
        "_data": {
          "activation_dtype": {
            "_data": "float8_e4m3fn",
            "_type": "torch.dtype"
          },
          "granularity": {
            "_data": {},
            "_type": "PerRow",
            "_version": 1
          },
          "mm_config": {
            "_data": {
              "emulate": false,
              "pad_inner_dim": false,
              "use_fast_accum": true
            },
            "_type": "Float8MMConfig",
            "_version": 1
          },
          "set_inductor_config": true,
          "weight_dtype": {
            "_data": "float8_e4m3fn",
            "_type": "torch.dtype"
          }
        },
        "_type": "Float8DynamicActivationFloat8WeightConfig",
        "_version": 1
      }
    },
    "quant_type_kwargs": {}
  },
  "resid_pdrop": 0.0,
  "rms_norm_eps": 1e-05,
  "rope_scaling": null,
  "rope_theta": 250000,
  "sliding_window": null,
  "tie_word_embeddings": false,
  "torch_dtype": "bfloat16",
  "transformers_version": "4.51.0.dev0",
  "use_cache": true,
  "vocab_size": 100352
}