Language Models are Super Mario: Absorbing Abilities from Homologous Models as a Free Lunch
Paper • 2311.03099 • Published • 35
How to use seed429/Teutonic-I-d3 with Transformers:
# Use a pipeline as a high-level helper
from transformers import pipeline
pipe = pipeline("text-generation", model="seed429/Teutonic-I-d3") # Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM
tokenizer = AutoTokenizer.from_pretrained("seed429/Teutonic-I-d3")
model = AutoModelForCausalLM.from_pretrained("seed429/Teutonic-I-d3")How to use seed429/Teutonic-I-d3 with vLLM:
# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "seed429/Teutonic-I-d3"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
-H "Content-Type: application/json" \
--data '{
"model": "seed429/Teutonic-I-d3",
"prompt": "Once upon a time,",
"max_tokens": 512,
"temperature": 0.5
}'docker model run hf.co/seed429/Teutonic-I-d3
How to use seed429/Teutonic-I-d3 with SGLang:
# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
--model-path "seed429/Teutonic-I-d3" \
--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": "seed429/Teutonic-I-d3",
"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 "seed429/Teutonic-I-d3" \
--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": "seed429/Teutonic-I-d3",
"prompt": "Once upon a time,",
"max_tokens": 512,
"temperature": 0.5
}'How to use seed429/Teutonic-I-d3 with Docker Model Runner:
docker model run hf.co/seed429/Teutonic-I-d3
This is a merge of pre-trained language models created using mergekit.
This model was merged using the DARE TIES merge method using downloaded_models/unconst_Teutonic-I-base3 as a base.
The following models were included in the merge:
The following YAML configuration was used to produce this model:
merge_method: dare_ties
base_model: downloaded_models/unconst_Teutonic-I-base3
dtype: bfloat16
parameters:
int_inv_lambda: 1.0
normalize: true
models:
- model: downloaded_models/best26_Teutonic-I-best2
parameters:
weight: 0.40
density: 0.90
- model: downloaded_models/ClarenceDan_Teutonic-I-Clarence-A5504
parameters:
weight: 0.33
density: 0.90
- model: downloaded_models/ChrisJackieChan_Teutonic-I-7
parameters:
weight: 0.27
density: 0.90