Author name cluster

Mehdi Rezagholizadeh

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

6 papers

2 author rows

NeurIPS Conference 2025 Conference Paper

Zebra-Llama: Towards Extremely Efficient Hybrid Models

Mingyu Yang
Mehdi Rezagholizadeh
Guihong Li
Vikram Appia
Emad Barsoum

With the growing demand for deploying large language models (LLMs) across diverse applications, improving their inference efficiency is crucial for sustainable and democratized access. However, retraining LLMs to meet new user-specific requirements is prohibitively expensive and environmentally unsustainable. In this work, we propose a practical and scalable alternative: composing efficient hybrid language models from existing pre-trained models. Our approach, X-EcoMLA, introduces a family of 1B, 3B, and 8B hybrid models by combining State Space Models (SSMs) and Multi-head Latent Attention (MLA) layers, using a refined initialization and post-training pipeline to efficiently transfer knowledge from pre-trained Transformers. X-EcoMLA achieves Transformer-level accuracy with near-SSM efficiency using only 7–11 billion training tokens (compared to the trillions required for pre-training) and an 8B teacher. Moreover, it dramatically reduces KV cache size—down to 3. 9%, 2%, and 2. 73% of the original for the 1B, 3B, and 8B variants, respectively—while preserving 100%, 100%, and over 97% of average zero-shot performance on LM Harness tasks. Compared to models like MambaInLLaMA, X-EcoMLA, Minitron, and Llamba, our approach consistently delivers competitive or superior accuracy while using significantly fewer tokens, smaller teachers, and vastly reduced KV cache memory. Notably, X-EcoMLA-8B surpasses Minitron-8B in few-shot accuracy by 7%, while using 8× fewer training tokens, over 12× smaller KV cache, and a smaller teacher (8B vs. 15B). It also achieves 1. 4x–3. 3x higher throughput (tokens/s) than MambaInLlama. The source code is released at https: //github. com/AMD-AGI/AMD-Hybrid-Models.

PDF Details

IJCAI Conference 2024 Conference Paper

Beyond the Limits: A Survey of Techniques to Extend the Context Length in Large Language Models

Xindi Wang
Mahsa Salmani
Parsa Omidi
Xiangyu Ren
Mehdi Rezagholizadeh
Armaghan Eshaghi

Recently, large language models (LLMs) have shown remarkable capabilities including understanding context, engaging in logical reasoning, and generating responses. However, this is achieved at the expense of stringent computational and memory requirements, hindering their ability to effectively support long input sequences. This survey provides an inclusive review of the recent techniques and methods devised to extend the sequence length in LLMs, thereby enhancing their capacity for long-context understanding. In particular, we review and categorize a wide range of techniques including architectural modifications, such as modified positional encoding and altered attention mechanisms, which are designed to enhance the processing of longer sequences while avoiding a proportional increase in computational cost. The diverse methodologies investigated in this study can be leveraged across different phases of LLMs, i. e. , training, fine-tuning and inference. This enables LLMs to efficiently process extended sequences. The limitations of the current methodologies is discussed in the last section along with the suggestions for future research directions, underscoring the importance of sequence length in the continued advancement of LLMs.

PDF Details DOI

AAAI Conference 2022 Conference Paper

From Fully Trained to Fully Random Embeddings: Improving Neural Machine Translation with Compact Word Embedding Tables

Krtin Kumar
Peyman Passban
Mehdi Rezagholizadeh
Yiusing Lau
Qun Liu

Embedding matrices are key components in neural natural language processing (NLP) models that are responsible to provide numerical representations of input tokens (i. e. words or subwords). In this paper, we analyze the impact and utility of such matrices in the context of neural machine translation (NMT). We show that detracting syntactic and semantic information from word embeddings and running NMT systems with random embeddings is not as damaging as it initially sounds. We also show how incorporating only a limited amount of taskspecific knowledge from fully-trained embeddings can boost the performance NMT systems. Our findings demonstrate that in exchange for negligible deterioration in performance, any NMT model can be run with partially random embeddings. Working with such structures means a minimal memory requirement as there is no longer need to store large embedding tables, which is a significant gain in industrial and on-device settings. We evaluated our embeddings in translating English into German and French and achieved a 5. 3x compression rate. Despite having a considerably smaller architecture, our models in some cases are even able to outperform state-of-the-art baselines.

PDF Details

UAI Conference 2022 Conference Paper

Learning functions on multiple sets using multi-set transformers

Kira A. Selby
Ahmad Rashid
Ivan Kobyzev
Mehdi Rezagholizadeh
Pascal Poupart

We propose a general deep architecture for learning functions on multiple permutation-invariant sets. We also show how to generalize this architecture to sets of elements of any dimension by dimension equivariance. We demonstrate that our architecture is a universal approximator of these functions, and show superior results to existing methods on a variety of tasks including counting tasks, alignment tasks, distinguishability tasks and statistical distance measurements. This last task is quite important in Machine Learning. Although our approach is quite general, we demonstrate that it can generate approximate estimates of KL divergence and mutual information that are more accurate than previous techniques that are specifically designed to approximate those statistical distances.

Details

AAAI Conference 2021 Conference Paper

ALP-KD: Attention-Based Layer Projection for Knowledge Distillation

Peyman Passban
Yimeng Wu
Mehdi Rezagholizadeh
Qun Liu

Knowledge distillation is considered as a training and compression strategy in which two neural networks, namely a teacher and a student, are coupled together during training. The teacher network is supposed to be a trustworthy predictor and the student tries to mimic its predictions. Usually, a student with a lighter architecture is selected so we can achieve compression and yet deliver high-quality results. In such a setting, distillation only happens for final predictions whereas the student could also benefit from teacher’s supervision for internal components. Motivated by this, we studied the problem of distillation for intermediate layers. Since there might not be a one-to-one alignment between student and teacher layers, existing techniques skip some teacher layers and only distill from a subset of them. This shortcoming directly impacts quality, so we instead propose a combinatorial technique which relies on attention. Our model fuses teacher-side information and takes each layer’s significance into consideration, then performs distillation between combined teacher layers and those of the student. Using our technique, we distilled a 12-layer BERT (Devlin et al. 2019) into 6-, 4-, and 2-layer counterparts and evaluated them on GLUE tasks (Wang et al. 2018). Experimental results show that our combinatorial approach is able to outperform other existing techniques.

PDF Details

NeurIPS Conference 2021 Conference Paper

NATURE: Natural Auxiliary Text Utterances for Realistic Spoken Language Evaluation

David Alfonso-Hermelo
Ahmad Rashid
Abbas Ghaddar
Philippe Langlais
Mehdi Rezagholizadeh

Slot-filling and intent detection are the backbone of conversational agents such as voice assistants, and are active areas of research. Even though state-of-the-art techniques on publicly available benchmarks show impressive performance, their ability to generalize to realistic scenarios is yet to be demonstrated. In this work, we present NATURE, a set of simple spoken-language-oriented transformations, applied to the evaluation set of datasets, to introduce human spoken language variations while preserving the semantics of an utterance. We apply NATURE to common slot-filling and intent detection benchmarks and demonstrate that simple perturbations from the standard evaluation set by NATURE can deteriorate model performance significantly. Through our experiments we demonstrate that when NATURE operators are applied to evaluation set of popular benchmarks the model accuracy can drop by up to 40%.

PDF Details