Resolving Interference When Merging Models
Paper • 2306.01708 • Published • 19
How to use Abin7/finnish-mal with Transformers:
# Use a pipeline as a high-level helper
from transformers import pipeline
pipe = pipeline("text-generation", model="Abin7/finnish-mal") # Load model directly
from transformers import AutoTokenizer, AutoModelForMultimodalLM
tokenizer = AutoTokenizer.from_pretrained("Abin7/finnish-mal")
model = AutoModelForMultimodalLM.from_pretrained("Abin7/finnish-mal")How to use Abin7/finnish-mal with vLLM:
# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Abin7/finnish-mal"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
-H "Content-Type: application/json" \
--data '{
"model": "Abin7/finnish-mal",
"prompt": "Once upon a time,",
"max_tokens": 512,
"temperature": 0.5
}'docker model run hf.co/Abin7/finnish-mal
How to use Abin7/finnish-mal with SGLang:
# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
--model-path "Abin7/finnish-mal" \
--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": "Abin7/finnish-mal",
"prompt": "Once upon a time,",
"max_tokens": 512,
"temperature": 0.5
}'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 "Abin7/finnish-mal" \
--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": "Abin7/finnish-mal",
"prompt": "Once upon a time,",
"max_tokens": 512,
"temperature": 0.5
}'How to use Abin7/finnish-mal with Docker Model Runner:
docker model run hf.co/Abin7/finnish-mal
This is a merge of pre-trained language models created using mergekit.
This model was merged using the TIES merge method using NousResearch/Llama-2-7b-hf as a base.
The following models were included in the merge:
The following YAML configuration was used to produce this model:
models:
- model: abhinand/malayalam-llama-7b-instruct-v0.1
parameters:
density: 0.5
weight: 0.5
- model: Finnish-NLP/llama-7b-finnish-instruct-v0.2
parameters:
density: 0.5
weight: 0.5
merge_method: ties
base_model: NousResearch/Llama-2-7b-hf
parameters:
normalize: false
int8_mask: true
dtype: float16