Instructions to use kazuHF/llm-jp-3-13b-it2_lora with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use kazuHF/llm-jp-3-13b-it2_lora with Transformers:

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

pipe = pipeline("text-generation", model="kazuHF/llm-jp-3-13b-it2_lora")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("kazuHF/llm-jp-3-13b-it2_lora")
model = AutoModelForCausalLM.from_pretrained("kazuHF/llm-jp-3-13b-it2_lora")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use kazuHF/llm-jp-3-13b-it2_lora with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "kazuHF/llm-jp-3-13b-it2_lora"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "kazuHF/llm-jp-3-13b-it2_lora",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/kazuHF/llm-jp-3-13b-it2_lora

SGLang

How to use kazuHF/llm-jp-3-13b-it2_lora 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 "kazuHF/llm-jp-3-13b-it2_lora" \
    --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": "kazuHF/llm-jp-3-13b-it2_lora",
		"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 "kazuHF/llm-jp-3-13b-it2_lora" \
        --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": "kazuHF/llm-jp-3-13b-it2_lora",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Unsloth Studio

How to use kazuHF/llm-jp-3-13b-it2_lora with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for kazuHF/llm-jp-3-13b-it2_lora to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for kazuHF/llm-jp-3-13b-it2_lora to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for kazuHF/llm-jp-3-13b-it2_lora to start chatting

Load model with FastModel

pip install unsloth
from unsloth import FastModel
model, tokenizer = FastModel.from_pretrained(
    model_name="kazuHF/llm-jp-3-13b-it2_lora",
    max_seq_length=2048,
)

Docker Model Runner
How to use kazuHF/llm-jp-3-13b-it2_lora with Docker Model Runner:
```
docker model run hf.co/kazuHF/llm-jp-3-13b-it2_lora
```

kazuHF commited on Dec 24, 2024

Commit

153984d

verified ·

1 Parent(s): bbc800e

Update README.md

Browse files

Files changed (1) hide show

README.md +6 -2

README.md CHANGED Viewed

@@ -66,14 +66,15 @@ model, tokenizer = FastLanguageModel.from_pretrained(
 # 元のモデルにLoRAのアダプタを統合。
 model = PeftModel.from_pretrained(model, adapter_id, token = HF_TOKEN)
-# 推論するためにモデルのモードを変更
-FastLanguageModel.for_inference(model)
 ```
 3. 単一の入力文に対して推論する場合
 ```bash
 # 単一の入力文に基づいて推論する関数の定義。
 def Decoder(input):
     prompt = f"""### 指示\n\n{str(input)}\n\n### 回答"""
     inputs = tokenizer([prompt], return_tensors = "pt").to(model.device)
     outputs = model.generate(**inputs, max_new_tokens = 512, use_cache = True, do_sample=False, repetition_penalty=1.2)
@@ -104,6 +105,9 @@ with open("./elyza-tasks-100-TV_0.jsonl", "r") as f:
 # モデルで入力を一括処理。
 results = []
 for dt in tqdm(datasets):
     input = dt["input"]
     prompt = f"""### 指示\n{input}\n### 回答\n"""
     inputs = tokenizer([prompt], return_tensors = "pt").to(model.device)

 # 元のモデルにLoRAのアダプタを統合。
 model = PeftModel.from_pretrained(model, adapter_id, token = HF_TOKEN)
 ```
 3. 単一の入力文に対して推論する場合
 ```bash
 # 単一の入力文に基づいて推論する関数の定義。
 def Decoder(input):
+    # 推論するためにモデルのモードを変更
+    FastLanguageModel.for_inference(model)
+    # 入力文による推論
     prompt = f"""### 指示\n\n{str(input)}\n\n### 回答"""
     inputs = tokenizer([prompt], return_tensors = "pt").to(model.device)
     outputs = model.generate(**inputs, max_new_tokens = 512, use_cache = True, do_sample=False, repetition_penalty=1.2)
 # モデルで入力を一括処理。
 results = []
 for dt in tqdm(datasets):
+    # 推論するためにモデルのモードを変更
+    FastLanguageModel.for_inference(model)
+    # 入力文による推論
     input = dt["input"]
     prompt = f"""### 指示\n{input}\n### 回答\n"""
     inputs = tokenizer([prompt], return_tensors = "pt").to(model.device)