Instructions to use DS-Archive/Nous-Hermes-Llama2-13b-Limarp-Lora-Merged with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use DS-Archive/Nous-Hermes-Llama2-13b-Limarp-Lora-Merged with Transformers:

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

pipe = pipeline("text-generation", model="DS-Archive/Nous-Hermes-Llama2-13b-Limarp-Lora-Merged")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("DS-Archive/Nous-Hermes-Llama2-13b-Limarp-Lora-Merged")
model = AutoModelForCausalLM.from_pretrained("DS-Archive/Nous-Hermes-Llama2-13b-Limarp-Lora-Merged")

Inference
Local Apps Settings

vLLM

How to use DS-Archive/Nous-Hermes-Llama2-13b-Limarp-Lora-Merged with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "DS-Archive/Nous-Hermes-Llama2-13b-Limarp-Lora-Merged"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "DS-Archive/Nous-Hermes-Llama2-13b-Limarp-Lora-Merged",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/DS-Archive/Nous-Hermes-Llama2-13b-Limarp-Lora-Merged

SGLang

How to use DS-Archive/Nous-Hermes-Llama2-13b-Limarp-Lora-Merged 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 "DS-Archive/Nous-Hermes-Llama2-13b-Limarp-Lora-Merged" \
    --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": "DS-Archive/Nous-Hermes-Llama2-13b-Limarp-Lora-Merged",
		"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 "DS-Archive/Nous-Hermes-Llama2-13b-Limarp-Lora-Merged" \
        --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": "DS-Archive/Nous-Hermes-Llama2-13b-Limarp-Lora-Merged",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use DS-Archive/Nous-Hermes-Llama2-13b-Limarp-Lora-Merged with Docker Model Runner:
```
docker model run hf.co/DS-Archive/Nous-Hermes-Llama2-13b-Limarp-Lora-Merged
```

Doctor-Shotgun commited on Jul 29, 2023

Commit

398335c

1 Parent(s): ef4ad7c

Create README.md

Browse files

Files changed (1) hide show

README.md +41 -0

README.md ADDED Viewed

	@@ -0,0 +1,41 @@

+---
+language:
+- en
+library_name: transformers
+pipeline_tag: text-generation
+tags:
+- llama-2
+---
+# Model Card: Nous-Hermes-Llama-2-13b-LIMARP-Lora-Merged
+This is Nous Hermes Llama 2 13b (https://huggingface.co/NousResearch/Nous-Hermes-Llama2-13b) merged with LIMARP Lora (https://huggingface.co/lemonilia/limarp-llama2) using the now-updated standard lora adapter for LIMARP (July 28, 2023).
+The intended objective was to combine NH-L2's reasoning and instruction-following capabilities with LIMARP's character roleplay capabilities.
+added_tokens.json was padded with dummy tokens to reach 32 added tokens in order to allow GGML conversion in llama.cpp without error due to vocab size mismatch.
+## Usage:
+Intended to be prompted either with Alpaca instruction format of the NH-L2 base model:
+```
+### Instruction:
+<prompt>
+### Response:
+<leave a newline blank for model to respond>
+```
+Or the LIMARP lora instruction format:
+```
+<<SYSTEM>>
+<character card and system prompt>
+<<USER>>
+<prompt>
+<<AIBOT>>
+<leave a newline blank for model to respond>
+```