CoRM-182M-top2

CoRM-182M (Top-2)

Contrastive Routing Mixture-of-Experts (CoRM). Checkpoint for the paper Beyond Magnitude: Contrastive Routing for Modular Mixture-of-Experts.

Part of the ilsp/CoRM collection.

Architecture

Usage

This model uses custom modeling code, so trust_remote_code=True is required.

from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "ilsp/CoRM-182M-top2"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, trust_remote_code=True)

inputs = tokenizer("The capital of Greece is", return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=32)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Training

<!-- TODO: fill in training data, token count, hardware, and hyperparameters. -->

Evaluation

<!-- TODO: add evaluation results. -->

Citation

<!-- TODO: add paper link and BibTeX. -->