Jithin VG

Adarsh MS, Jithin VG, Ditto PS

Large language models (LLMs) are known for their exceptional performance across a range of natural language processing tasks, but their deployment comes at a high computational and financial cost. On the other hand, smaller language models (SLMs), which can be deployed on lower-cost edge devices, struggle to match the performance of their larger counterparts. This paper presents a novel hybrid inference approach that leverages the strengths of both model types while minimizing reliance on costly cloud-based LLMs. Unlike existing methods that route entire queries to either an SLM or a cloud LLM, our approach introduces a reward-based mechanism to dynamically determine the involvement of the cloud LLM during token generation. Specifically, each token predicted by the SLM is evaluated against a reward score, and only when this score falls below a certain threshold is the cloud LLM consulted for assistance in the next token prediction. This method not only reduces the traffic to the cloud LLM, thereby lowering costs, but also allows for flexible control over response quality depending on the reward score threshold. Experimental results demonstrate that our approach significantly reduces cloud LLM usage with minimal impact on overall response quality, offering a cost-effective solution for deploying high-performance language models

Ditto PS, Jithin VG, Adarsh MS

In recent years, large language models have demonstrated remarkable performance across various natural language processing (NLP) tasks. However, deploying these models for real-world applications often requires efficient inference solutions to handle the computational demands. In this paper, we explore the utilization of CPUs for accelerating the inference of large language models. Specifically, we introduce a parallelized approach to enhance throughput by 1) Exploiting the parallel processing capabilities of modern CPU architectures, 2) Batching the inference request. Our evaluation shows the accelerated inference engine gives an 18-22x improvement in the generated token per sec. The improvement is more with longer sequence and larger models. In addition to this, we can also run multiple workers in the same machine with NUMA node isolation to further improvement in tokens/s. Table 2, we have received 4x additional improvement with 4 workers. This would also make Gen-AI based products and companies environment friendly, our estimates shows that CPU usage for Inference could reduce the power consumption of LLMs by 48.9% while providing production ready throughput and latency.

Jithin VG, Ditto PS

The proliferation of GPU-accelerated workloads, particularly in artificial intelligence and large language model (LLM) inference, has created unprecedented demand for efficient GPU resource sharing in cloud and container environments. While NVIDIA's Multi-Instance GPU (MIG) technology provides hardware-level isolation, its availability is limited to high-end datacenter GPUs. Software-based virtualization solutions such as HAMi-core and BUD-FCSP offer alternatives for broader GPU families but lack standardized evaluation methodologies. We present GPU-Virt-Bench, a comprehensive benchmarking framework that evaluates GPU virtualization systems across 56 performance metrics organized into 10 categories. Our framework measures overhead, isolation quality, LLM-specific performance, memory bandwidth, cache behavior, PCIe throughput, multi-GPU communication, scheduling efficiency, memory fragmentation, and error recovery. GPU-Virt-Bench enables systematic comparison between software virtualization approaches and ideal MIG behavior, providing actionable insights for practitioners deploying GPU resources in multi-tenant environments. We demonstrate the framework's utility through evaluation of HAMi-core, BUD-FCSP, and simulated MIG baselines, revealing performance characteristics critical for production deployment decisions.

Ajmal PS, Ditto PS, Jithin VG

Intellecta dataset emerges as an innovative synthetic dataset, engineered to enhance the cognitive processing capabilities of contemporary language models. With a composition of 11.53 billion tokens, integrating 8.01 billion tokens of synthetic data with 3.52 billion tokens of rich textbook data, Intellecta is crafted to foster advanced reasoning and comprehensive educational narrative generation. Leveraging the Mixtral-8x7B-Instruct-v0.1 model, the dataset facilitates the generation of complex thought processes and detailed, textbook-style explanations, thus enabling language models to engage in both critical thinking and profound educational discourse. This hybrid dataset stands as a testament to the potential of synthetic data in pushing the boundaries of AI, offering a repository that is not only vast and varied but also refined to align with ethical standards and intellectual rigor.