Instructions to use whynlp/pccot-llama1b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use whynlp/pccot-llama1b with Transformers:

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

pipe = pipeline("text-generation", model="whynlp/pccot-llama1b")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("whynlp/pccot-llama1b", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use whynlp/pccot-llama1b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "whynlp/pccot-llama1b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "whynlp/pccot-llama1b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/whynlp/pccot-llama1b

SGLang

How to use whynlp/pccot-llama1b 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 "whynlp/pccot-llama1b" \
    --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": "whynlp/pccot-llama1b",
		"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 "whynlp/pccot-llama1b" \
        --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": "whynlp/pccot-llama1b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use whynlp/pccot-llama1b with Docker Model Runner:
```
docker model run hf.co/whynlp/pccot-llama1b
```

whynlp commited on Mar 29

Commit

9f23cce

verified ·

1 Parent(s): f7a9012

Update README.md

Browse files

Files changed (1) hide show

README.md +6 -5

README.md CHANGED Viewed

@@ -93,8 +93,9 @@ Disclaimer: The weights of this model is not the original weights of the model t
 The model is finetuned on the [GSM8K-AUG](https://huggingface.co/datasets/whyNLP/gsm8k-aug) dataset, achieving the following performance:
-| Model                           | Accuracy             |
-| ------------------------------- | -------------------- |
-| CoT (llama1b)                   | 61.6                 |
-| [PCCoT-gpt2](whynlp/pccot-gpt2) | 49.48 \\(\pm\\) 0.31 |
-| **PCCoT-llama1b**               | 53.35 \\(\pm\\) 0.18 |

 The model is finetuned on the [GSM8K-AUG](https://huggingface.co/datasets/whyNLP/gsm8k-aug) dataset, achieving the following performance:
+| Model                                                        | Accuracy             |
+| ------------------------------------------------------------ | -------------------- |
+| CoT (llama1b)                                                | 61.6                 |
+| [PCCoT-gpt2](https://huggingface.co/whynlp/pccot-gpt2)       | 49.48 \\(\pm\\) 0.31 |
+| [PCCoT-gpt2-nl](https://huggingface.co/whynlp/pccot-gpt2-nl) | 49.23 \\(\pm\\) 0.80 |
+| **PCCoT-llama1b**                                            | 53.35 \\(\pm\\) 0.18 |