Instructions to use DukeNLP/Prob-Gen-8B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use DukeNLP/Prob-Gen-8B with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="DukeNLP/Prob-Gen-8B")# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("DukeNLP/Prob-Gen-8B") model = AutoModelForCausalLM.from_pretrained("DukeNLP/Prob-Gen-8B") - Notebooks
- Google Colab
- Kaggle
- Local Apps
- vLLM
How to use DukeNLP/Prob-Gen-8B with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "DukeNLP/Prob-Gen-8B" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "DukeNLP/Prob-Gen-8B", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/DukeNLP/Prob-Gen-8B
- SGLang
How to use DukeNLP/Prob-Gen-8B 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 "DukeNLP/Prob-Gen-8B" \ --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": "DukeNLP/Prob-Gen-8B", "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 "DukeNLP/Prob-Gen-8B" \ --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": "DukeNLP/Prob-Gen-8B", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use DukeNLP/Prob-Gen-8B with Docker Model Runner:
docker model run hf.co/DukeNLP/Prob-Gen-8B
Update README.md
Browse files
README.md
CHANGED
|
@@ -31,38 +31,21 @@ This model has been fine-tuned using 4-bit QLORA, based on [Llama-3-8B from Meta
|
|
| 31 |
|
| 32 |
The model can be loaded with HuggingFace's Transformers library:
|
| 33 |
``` python
|
| 34 |
-
|
| 35 |
-
import torch
|
| 36 |
|
| 37 |
-
model_id = "
|
| 38 |
|
| 39 |
-
model = AutoModelForCausalLM.from_pretrained(
|
| 40 |
-
model_id,
|
| 41 |
-
device_map="auto",
|
| 42 |
-
torch_dtype=torch.bfloat16
|
| 43 |
-
)
|
| 44 |
|
| 45 |
-
tokenizer = AutoTokenizer.from_pretrained(
|
| 46 |
-
model_id,
|
| 47 |
-
use_fast=False,
|
| 48 |
-
legacy=False
|
| 49 |
-
)
|
| 50 |
|
| 51 |
-
|
| 52 |
-
|
| 53 |
-
|
| 54 |
-
|
| 55 |
-
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
model_output = model.generate(
|
| 59 |
-
model_input['input_ids'],
|
| 60 |
-
max_new_tokens=256,
|
| 61 |
-
do_sample=True,
|
| 62 |
-
...
|
| 63 |
-
)
|
| 64 |
-
|
| 65 |
-
tokenizer.batch_decode(model_output)
|
| 66 |
```
|
| 67 |
|
| 68 |
<!-- ## Bias, Risks, and Limitations
|
|
|
|
| 31 |
|
| 32 |
The model can be loaded with HuggingFace's Transformers library:
|
| 33 |
``` python
|
| 34 |
+
from transformers import AutoModelForCausalLM, AutoTokenizer
|
|
|
|
| 35 |
|
| 36 |
+
model_id = "DukeNLP/Prob-Gen-8B"
|
| 37 |
|
| 38 |
+
model = AutoModelForCausalLM.from_pretrained(model_id,device_map="auto", trust_remote_code=True)
|
|
|
|
|
|
|
|
|
|
|
|
|
| 39 |
|
| 40 |
+
tokenizer = AutoTokenizer.from_pretrained(model_id)
|
|
|
|
|
|
|
|
|
|
|
|
|
| 41 |
|
| 42 |
+
prompt = "Please generate a math problem and 2 to 4 options for 8th graders with the following requirements:\nProblem context: <specified-context>\nTested knowledge: <specified-knowledge>"
|
| 43 |
+
|
| 44 |
+
model_input = tokenizer(prompt, return_tensors="pt").to("cuda")
|
| 45 |
+
|
| 46 |
+
model_output = model.generate(model_input['input_ids'], max_new_tokens=256)
|
| 47 |
+
|
| 48 |
+
print(tokenizer.batch_decode(model_output))
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 49 |
```
|
| 50 |
|
| 51 |
<!-- ## Bias, Risks, and Limitations
|