Instructions to use eyepyon/rcPhi4miniR-finetuned with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use eyepyon/rcPhi4miniR-finetuned with Transformers:

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

pipe = pipeline("text-generation", model="eyepyon/rcPhi4miniR-finetuned")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("eyepyon/rcPhi4miniR-finetuned", dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use eyepyon/rcPhi4miniR-finetuned with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "eyepyon/rcPhi4miniR-finetuned"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "eyepyon/rcPhi4miniR-finetuned",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/eyepyon/rcPhi4miniR-finetuned

SGLang

How to use eyepyon/rcPhi4miniR-finetuned 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 "eyepyon/rcPhi4miniR-finetuned" \
    --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": "eyepyon/rcPhi4miniR-finetuned",
		"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 "eyepyon/rcPhi4miniR-finetuned" \
        --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": "eyepyon/rcPhi4miniR-finetuned",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use eyepyon/rcPhi4miniR-finetuned with Docker Model Runner:
```
docker model run hf.co/eyepyon/rcPhi4miniR-finetuned
```

eyepyon/rcPhi4miniR-finetuned

このモデルは、microsoft/Phi-4-mini-reasoning をベースにLoRAでファインチューニングされたモデルです。

モデル情報

ベースモデル: microsoft/Phi-4-mini-reasoning
ファインチューニング手法: LoRA (Low-Rank Adaptation)
アテンション実装: eager (Phi-4推奨)
量子化: 4ビット (QLoRA)
対応言語: 日本語
タスク: 質問応答

使用方法

from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel

# ベースモデルとトークナイザーを読み込み
base_model = AutoModelForCausalLM.from_pretrained(
    "microsoft/Phi-4-mini-reasoning",
    torch_dtype="auto",
    device_map="auto",
    attn_implementation="eager",
    trust_remote_code=True
)
tokenizer = AutoTokenizer.from_pretrained("microsoft/Phi-4-mini-reasoning")

# LoRAアダプターを適用
model = PeftModel.from_pretrained(base_model, "eyepyon/rcPhi4miniR-finetuned")

# 推論
input_text = "### コンテキスト:\n[コンテキスト]\n\n### 質問:\n[質問]\n\n### 回答:\n"
inputs = tokenizer(input_text, return_tensors="pt")
outputs = model.generate(**inputs, max_length=200, do_sample=True, temperature=0.7)
response = tokenizer.decode(outputs[0], skip_special_tokens=True)
print(response)

入力フォーマット

### コンテキスト:
[背景情報]

### 質問:
[ユーザーの質問]

### 回答:
[期待される回答]

トレーニング設定

LoRAランク: 8
LoRA Alpha: 16
ターゲットモジュール: qkv_proj, o_proj, gate_up_proj, down_proj
学習率: 2e-4
バッチサイズ: 1 × 4 (gradient accumulation)
エポック数: 3

注意事項

Phi-4モデル使用のため attn_implementation="eager" が推奨されます
use_cache=False でトレーニングされています
グラディエントチェックポイント対応済み

Downloads last month: -; Downloads are not tracked for this model. How to track

Model tree for eyepyon/rcPhi4miniR-finetuned

Base model

microsoft/Phi-4-mini-reasoning

Adapter

(7)

this model