roneneldan/TinyStories
Viewer • Updated • 2.14M • 91.4k • 1k
How to use doabell/olmostories-29m with Transformers:
# Use a pipeline as a high-level helper
from transformers import pipeline
pipe = pipeline("text-generation", model="doabell/olmostories-29m") # Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM
tokenizer = AutoTokenizer.from_pretrained("doabell/olmostories-29m")
model = AutoModelForCausalLM.from_pretrained("doabell/olmostories-29m")How to use doabell/olmostories-29m with vLLM:
# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "doabell/olmostories-29m"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
-H "Content-Type: application/json" \
--data '{
"model": "doabell/olmostories-29m",
"prompt": "Once upon a time,",
"max_tokens": 512,
"temperature": 0.5
}'docker model run hf.co/doabell/olmostories-29m
How to use doabell/olmostories-29m with SGLang:
# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
--model-path "doabell/olmostories-29m" \
--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": "doabell/olmostories-29m",
"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 "doabell/olmostories-29m" \
--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": "doabell/olmostories-29m",
"prompt": "Once upon a time,",
"max_tokens": 512,
"temperature": 0.5
}'How to use doabell/olmostories-29m with Docker Model Runner:
docker model run hf.co/doabell/olmostories-29m
TinyStories trained on OLMo 2 architecture.
Took around 4 hours on an A100 (80GB).
config = Olmo2Config(
vocab_size=5000,
hidden_size=512,
intermediate_size=1280,
num_hidden_layers=8,
num_attention_heads=8,
num_key_value_heads=8,
max_position_embeddings=1024,
initializer_range=0.02,
attention_dropout=0.1,
)