Instructions to use ffurfaro/Titans-OpenELM-1_1B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ffurfaro/Titans-OpenELM-1_1B with Transformers:

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

pipe = pipeline("text-generation", model="ffurfaro/Titans-OpenELM-1_1B")

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("ffurfaro/Titans-OpenELM-1_1B", device_map="auto")

PEFT
How to use ffurfaro/Titans-OpenELM-1_1B with PEFT:
```
Task type is invalid.
```
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use ffurfaro/Titans-OpenELM-1_1B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ffurfaro/Titans-OpenELM-1_1B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ffurfaro/Titans-OpenELM-1_1B",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/ffurfaro/Titans-OpenELM-1_1B

SGLang

How to use ffurfaro/Titans-OpenELM-1_1B 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 "ffurfaro/Titans-OpenELM-1_1B" \
    --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": "ffurfaro/Titans-OpenELM-1_1B",
		"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 "ffurfaro/Titans-OpenELM-1_1B" \
        --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": "ffurfaro/Titans-OpenELM-1_1B",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use ffurfaro/Titans-OpenELM-1_1B with Docker Model Runner:
```
docker model run hf.co/ffurfaro/Titans-OpenELM-1_1B
```

Titans-OpenELM-1_1B / README.md

ffurfaro

Upload folder using huggingface_hub

f4eeddd verified about 1 year ago

preview code

Raw History Blame Contribute Delete

3.18 kB

---
language: en
license: apache-2.0
library_name: transformers
tags:
  - tptt
  - peft
  - trust_remote_code
pipeline_tag: text-generation
base_model: apple/OpenELM-1_1B
datasets:
- yahma/alpaca-cleaned
---
# Titanesque-OpenELM-1_1B
<p align="center">
    <a href="https://arxiv.org/abs/2506.17671">
        <img alt="arXiv" src="https://img.shields.io/badge/arXiv-tptt-blueviolet.svg">
    </a>
    <a href="https://pypi.org/project/tptt/">
        <img alt="PyPI" src="https://img.shields.io/pypi/v/tptt?color=orange">
    </a>
    <a href="https://github.com/fabienfrfr/tptt/">
        <img alt="Release" src="https://img.shields.io/github/v/release/fabienfrfr/tptt?color=brightgreen">
    </a>
    <a href="https://fabienfrfr.github.io/tptt/">
        <img alt="Documentation" src="https://img.shields.io/badge/docs-online-blue">
    </a>
    <a href="https://huggingface.co/ffurfaro">
        <img alt="HuggingFace" src="https://img.shields.io/badge/hf-ffurfaro-yellow">
    </a>
</p>
Titanesque version of `apple/OpenELM-1_1B` with parallel linearized attention (TPTT 😊) and PEFT.

The architecture was presented in the paper [TPTT](https://huggingface.co/papers/2506.17671).


## Model list

Classic model parameter with LiZA injection :

| Subfolder                      | Max Self Attn Length | Mag Weight | Cross Gate | Max Chunk Size | Bidirectional | LoRA | Description                                           |
|-------------------------------|----------------------|------------|------------|----------------|---------------|------|-------------------------------------------------------|
| delta_rule  | 8192 (default)       | 0.5        | False      | 64             | False         | Yes  | Parallel linearized attention with delta_rule operator|
| delta_rule_gelu | 8192 (default) | 0.5        | False      | 64             | False         | Yes  | Non-linear operator with gelu activation              |
| delta_product    | 8192 (default) | 0.5        | False      | 64             | False         | Yes  | Second order operator with derivative trick              |
| delta_product_r  | 8192 (default) | 0.5        | False      | 64             | False         | Yes  | Second order operator with rotative trick             |
| delta_product_c  | 8192 (default) | 0.5        | False      | 64             | False         | Yes  | Second order operator with combined trick             |
## Usage
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
model = AutoModelForCausalLM.from_pretrained(
"ffurfaro/Titanesque-OpenELM-1_1B",
subfolder="tptt_subfolder", # see in repo tree
trust_remote_code=True
)
tokenizer = AutoTokenizer.from_pretrained("ffurfaro/apple/OpenELM-1_1B")

prompt = "Your prompt here"
inputs = tokenizer(prompt, return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=100)
print(tokenizer.decode(outputs, skip_special_tokens=True))
```
## Citation & Contact
If you use TPTT in your academic work, please cite [Furfaro](https://huggingface.co/ffurfaro). For questions or support, please open an issue on the [GitHub repository](https://github.com/fabienfrfr/tptt) or contact the maintainer.
---

	---
	language: en
	license: apache-2.0
	library_name: transformers
	tags:
	- tptt
	- peft
	- trust_remote_code
	pipeline_tag: text-generation
	base_model: apple/OpenELM-1_1B
	datasets:
	- yahma/alpaca-cleaned
	---

	# Titanesque-OpenELM-1_1B

	<p align="center">
	<a href="https://arxiv.org/abs/2506.17671">
	<img alt="arXiv" src="https://img.shields.io/badge/arXiv-tptt-blueviolet.svg">
	</a>
	<a href="https://pypi.org/project/tptt/">
	<img alt="PyPI" src="https://img.shields.io/pypi/v/tptt?color=orange">
	</a>
	<a href="https://github.com/fabienfrfr/tptt/">
	<img alt="Release" src="https://img.shields.io/github/v/release/fabienfrfr/tptt?color=brightgreen">
	</a>
	<a href="https://fabienfrfr.github.io/tptt/">
	<img alt="Documentation" src="https://img.shields.io/badge/docs-online-blue">
	</a>
	<a href="https://huggingface.co/ffurfaro">
	<img alt="HuggingFace" src="https://img.shields.io/badge/hf-ffurfaro-yellow">
	</a>
	</p>

	Titanesque version of `apple/OpenELM-1_1B` with parallel linearized attention (TPTT 😊) and PEFT.

	The architecture was presented in the paper [TPTT](https://huggingface.co/papers/2506.17671).


	## Model list

	Classic model parameter with LiZA injection :

	\| Subfolder \| Max Self Attn Length \| Mag Weight \| Cross Gate \| Max Chunk Size \| Bidirectional \| LoRA \| Description \|
	\|-------------------------------\|----------------------\|------------\|------------\|----------------\|---------------\|------\|-------------------------------------------------------\|
	\| delta_rule \| 8192 (default) \| 0.5 \| False \| 64 \| False \| Yes \| Parallel linearized attention with delta_rule operator\|
	\| delta_rule_gelu \| 8192 (default) \| 0.5 \| False \| 64 \| False \| Yes \| Non-linear operator with gelu activation \|
	\| delta_product \| 8192 (default) \| 0.5 \| False \| 64 \| False \| Yes \| Second order operator with derivative trick \|
	\| delta_product_r \| 8192 (default) \| 0.5 \| False \| 64 \| False \| Yes \| Second order operator with rotative trick \|
	\| delta_product_c \| 8192 (default) \| 0.5 \| False \| 64 \| False \| Yes \| Second order operator with combined trick \|

	## Usage

	```python
	from transformers import AutoModelForCausalLM, AutoTokenizer

	model = AutoModelForCausalLM.from_pretrained(
	"ffurfaro/Titanesque-OpenELM-1_1B",
	subfolder="tptt_subfolder", # see in repo tree
	trust_remote_code=True
	)
	tokenizer = AutoTokenizer.from_pretrained("ffurfaro/apple/OpenELM-1_1B")

	prompt = "Your prompt here"
	inputs = tokenizer(prompt, return_tensors="pt")
	outputs = model.generate(**inputs, max_new_tokens=100)
	print(tokenizer.decode(outputs, skip_special_tokens=True))

	```


	## Citation & Contact

	If you use TPTT in your academic work, please cite [Furfaro](https://huggingface.co/ffurfaro). For questions or support, please open an issue on the [GitHub repository](https://github.com/fabienfrfr/tptt) or contact the maintainer.


	---