Titans-OLMo-1B-hf

Titanesque version of allenai/OLMo-1B-hf with parallel linearized attention (TPTT 😊) and PEFT.

The model was presented in the paper TPTT.

For code, see https://github.com/fabienfrfr/tptt

Model Details

• Architecture: TpttModel
• Base model: allenai/OLMo-1B-hf
• LiZA config: operator=delta_rule, mag=0.5
• LoRA config: r=8, alpha=16, dropout=0.05
• torch_dtype: bfloat16

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained(
"ffurfaro/Titans-OLMo-1B-hf",
trust_remote_code=True
)
tokenizer = AutoTokenizer.from_pretrained("ffurfaro/Titans-OLMo-1B-hf")

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))

Training

• Dataset: yahma/alpaca-cleaned
• Platform: Kaggle
• Hardware: NVIDIA 2xT4
• Batch size: 3
• Epochs: 5.0
• Learning rate (final): 1.1904761904761906e-06
• Loss (final): 1.3068
• Training runtime: 1585.151 sec
• Samples per second: 1.577
• Steps per second: 0.265
• Total FLOPs: 6196832501760000.0
• Gradient norm (final): 3.117034673690796

Evaluation

• Metrics: Training loss only (no eval yet, table soon : PiQA, ARC, Hella, Wino, GSM8K, MMLU)
• Results: Final training loss: 1.3068

Citation & Contact

If you use TPTT in your academic work, please cite Furfaro. For questions or support, please open an issue on the GitHub repository or contact the maintainer.

---