Model Stock: All we need is just a few fine-tuned models
Paper • 2403.19522 • Published • 15
How to use ehristoforu/0001 with Transformers:
# Use a pipeline as a high-level helper
from transformers import pipeline
pipe = pipeline("text-generation", model="ehristoforu/0001") # Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM
tokenizer = AutoTokenizer.from_pretrained("ehristoforu/0001")
model = AutoModelForCausalLM.from_pretrained("ehristoforu/0001")How to use ehristoforu/0001 with vLLM:
# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ehristoforu/0001"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
-H "Content-Type: application/json" \
--data '{
"model": "ehristoforu/0001",
"prompt": "Once upon a time,",
"max_tokens": 512,
"temperature": 0.5
}'docker model run hf.co/ehristoforu/0001
How to use ehristoforu/0001 with SGLang:
# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
--model-path "ehristoforu/0001" \
--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": "ehristoforu/0001",
"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 "ehristoforu/0001" \
--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": "ehristoforu/0001",
"prompt": "Once upon a time,",
"max_tokens": 512,
"temperature": 0.5
}'How to use ehristoforu/0001 with Docker Model Runner:
docker model run hf.co/ehristoforu/0001
This is a merge of pre-trained language models created using mergekit.
This model was merged using the Model Stock merge method using Gaivoronsky/Mistral-7B-Saiga as a base.
The following models were included in the merge:
The following YAML configuration was used to produce this model:
slices:
- sources:
- model: Gaivoronsky/Mistral-7B-Saiga
layer_range:
- 0
- 32
- model: HuggingFaceH4/mistral-7b-grok
layer_range:
- 0
- 32
- model: NousResearch/Yarn-Mistral-7b-128k
layer_range:
- 0
- 32
merge_method: model_stock
base_model: Gaivoronsky/Mistral-7B-Saiga
parameters:
t:
- filter: self_attn
value:
- 0
- 0.5
- 0.3
- 0.7
- 1
- filter: mlp
value:
- 1
- 0.5
- 0.7
- 0.3
- 0
- filter: mlp
value:
- 1
- 0.5
- 0.7
- 0.3
- 0
- value: 0.5
dtype: bfloat16