Instructions to use phanerozoic/PirateTalk-13b-v2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use phanerozoic/PirateTalk-13b-v2 with Transformers:

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

pipe = pipeline("text-generation", model="phanerozoic/PirateTalk-13b-v2")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("phanerozoic/PirateTalk-13b-v2")
model = AutoModelForCausalLM.from_pretrained("phanerozoic/PirateTalk-13b-v2", device_map="auto")

Inference
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use phanerozoic/PirateTalk-13b-v2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "phanerozoic/PirateTalk-13b-v2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "phanerozoic/PirateTalk-13b-v2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/phanerozoic/PirateTalk-13b-v2

SGLang

How to use phanerozoic/PirateTalk-13b-v2 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 "phanerozoic/PirateTalk-13b-v2" \
    --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": "phanerozoic/PirateTalk-13b-v2",
		"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 "phanerozoic/PirateTalk-13b-v2" \
        --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": "phanerozoic/PirateTalk-13b-v2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use phanerozoic/PirateTalk-13b-v2 with Docker Model Runner:
```
docker model run hf.co/phanerozoic/PirateTalk-13b-v2
```

This repository contains the PirateTalk-13b-v2 model, our second and slightly better attempt at imbuing Llama 2 13b with persistent and unprompted pirate vernacular.

Overview: The PirateTalk-13b-v2 repository exemplifies our dedication to exploring domain-specific dialects within language models. Building on the Llama 2 Chat architecture and informed by the MistralPirate project, this version epitomizes the integration of thematic vernacular.

Objective: In alignment with our research interests, this iteration aims to seamlessly integrate and enhance the intricacies of the pirate dialect within the model's framework.

Base Model: Although the Mistral architecture was a pivotal exploration point, PirateTalk-13b-v2 roots itself back into the Llama 2 13b Chat model, leveraging its strengths to further our research on thematic vernacular.

Dataset: Our core pirate-themed dataset is unchanged from MistralPirate and PirateTalk-v2. Entries consisted of plaintext lines with no formatting aside from newlines.

Performance Insights: PirateTalk-13b-v2 demonstrates improved response conciseness, a direct outcome of our experiences with MistralPirate. The linguistic depth remains, but with enhanced coherence.

Research Trajectories: Our focus remains on advancing domain-specific dialects within language models. Upcoming endeavors will delve deeper into fine-tuning techniques, dataset enhancements, and potential architectural synergies.

Downloads last month: 9