Instructions to use MoxoffSrL/Moxoff-Phi3Mini-KTO with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use MoxoffSrL/Moxoff-Phi3Mini-KTO with Transformers:

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

pipe = pipeline("text-generation", model="MoxoffSrL/Moxoff-Phi3Mini-KTO", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("MoxoffSrL/Moxoff-Phi3Mini-KTO", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("MoxoffSrL/Moxoff-Phi3Mini-KTO", trust_remote_code=True, 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=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use MoxoffSrL/Moxoff-Phi3Mini-KTO with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "MoxoffSrL/Moxoff-Phi3Mini-KTO"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MoxoffSrL/Moxoff-Phi3Mini-KTO",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/MoxoffSrL/Moxoff-Phi3Mini-KTO

SGLang

How to use MoxoffSrL/Moxoff-Phi3Mini-KTO 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 "MoxoffSrL/Moxoff-Phi3Mini-KTO" \
    --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": "MoxoffSrL/Moxoff-Phi3Mini-KTO",
		"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 "MoxoffSrL/Moxoff-Phi3Mini-KTO" \
        --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": "MoxoffSrL/Moxoff-Phi3Mini-KTO",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use MoxoffSrL/Moxoff-Phi3Mini-KTO with Docker Model Runner:
```
docker model run hf.co/MoxoffSrL/Moxoff-Phi3Mini-KTO
```

JacopoAbate commited on Jun 27, 2024

Commit

ef38f68

verified ·

1 Parent(s): 2eded3a

Update README.md

Browse files

Files changed (1) hide show

README.md +5 -5

README.md CHANGED Viewed

@@ -17,7 +17,7 @@ metrics:
 # Model Information
-Phi3-instruct-128k-KTO is an updated version of [Phi-3-mini-128k-instruct](https://huggingface.co/microsoft/Phi-3-mini-128k-instruct), aligned with KTO and QLora.
 - It's trained on [distilabel-intel-orca-kto](https://huggingface.co/datasets/argilla/distilabel-intel-orca-kto).
@@ -27,7 +27,7 @@ We evaluated the model using the same test sets as used for the [Open LLM Leader
 | hellaswag acc_norm | arc_challenge acc_norm | m_mmlu 5-shot acc | Average |
 |:----------------------| :--------------- | :-------------------- | :------- |
-| 0.7915 | 0.5606 | 0.6939 | 0.6357 |
 ## Usage
@@ -43,8 +43,8 @@ from transformers import AutoModelForCausalLM, AutoTokenizer
 device = "cpu" # if you want to use the gpu make sure to have cuda toolkit installed and change this to "cuda"
-model = AutoModelForCausalLM.from_pretrained("MoxoffSpA/Phi3-instruct-128k-KTO")
-tokenizer = AutoTokenizer.from_pretrained("MoxoffSpA/Phi3-instruct-128k-KTO")
 question = """Quanto è alta la torre di Pisa?"""
 context = """
@@ -78,7 +78,7 @@ print(trimmed_output)
 ## Bias, Risks and Limitations
-Phi3-instruct-128k-KTO has not been aligned to human preferences for safety within the RLHF phase or deployed with in-the-loop filtering of
 responses like ChatGPT, so the model can produce problematic outputs (especially when prompted to do so). It is also unknown what the size and composition
 of the corpus was used to train the base model, however it is likely to have included a mix of Web data and technical sources
 like books and code.

 # Model Information
+Moxoff-Phi3Mini-KTO is an updated version of [Phi-3-mini-128k-instruct](https://huggingface.co/microsoft/Phi-3-mini-128k-instruct), aligned with KTO and QLora.
 - It's trained on [distilabel-intel-orca-kto](https://huggingface.co/datasets/argilla/distilabel-intel-orca-kto).
 | hellaswag acc_norm | arc_challenge acc_norm | m_mmlu 5-shot acc | Average |
 |:----------------------| :--------------- | :-------------------- | :------- |
+| 0.7915 | 0.5606 | 0.6939 | 0.682 |
 ## Usage
 device = "cpu" # if you want to use the gpu make sure to have cuda toolkit installed and change this to "cuda"
+model = AutoModelForCausalLM.from_pretrained("MoxoffSpA/Moxoff-Phi3Mini-KTO")
+tokenizer = AutoTokenizer.from_pretrained("MoxoffSpA/Moxoff-Phi3Mini-KTO")
 question = """Quanto è alta la torre di Pisa?"""
 context = """
 ## Bias, Risks and Limitations
+Moxoff-Phi3Mini-KTO has not been aligned to human preferences for safety within the RLHF phase or deployed with in-the-loop filtering of
 responses like ChatGPT, so the model can produce problematic outputs (especially when prompted to do so). It is also unknown what the size and composition
 of the corpus was used to train the base model, however it is likely to have included a mix of Web data and technical sources
 like books and code.