The Chicago Journal

The Chicago Journal
The Chicago Journal

Lingjing(Ginny) Jiang: “Space” is the gap between the body and itself

The Chicago Journal Staff on October 25, 2023

“Tides” is an innovative dance film project with a female perspective at its core. It aims to shed light on the heightened vulnerability that women often experience under societal scrutiny. This artful endeavor involves the intricate folding of the human body into diverse spaces, serving as a powerful means of expression and resistance against insecurity and chronic oppression. A bunch of art festivals and exhibitions select this film. Maggie Allesee Choreography Award Competition， Dance Camera West, Indie Short Festival， Video Art and Experimental Film， Big Syn International Film Festival， Fort Lauderdale International Film Festival， Florence Contemporary Art Exchange Exhibition, Art week Italia di Santa Maria Del Fiore, InShadow-Lisbon Screendance Festival, London Temporal Collision, Sohu Short Film Award Ceremony, Shanghai Dance Festival, Xiamen Art Exhibition.

Interviewer: As a director, how did you balance the visual effects of the image, the rhythm of the music, and the performances of the dancers during the creation of “Tides”?

Ginny: In the initial thought, the concept and imagery over structure or choreography were settled. The music evolved through various versions during editing, with the final choice offering non-order and ample expressive room. I preset the movement, but much remained improvised to avoid limiting the dancer’s freedom. This approach allowed me to both disrupt and maintain elements of dance coherence. Dance and experimental imagery and film composition should transcend a dance performance. The body serves as a medium for expression rather than a stiff storytelling or performance.

Interviewer: As a performer, how do you use your body language to convey the relationship with nature, individuality, and space? Were these physical movements pre-planned or improvised?

Ginny: The body, for us as dance artists, is our tool; It’s the medium through which I express “myself.” Space, to me, is the gap between my body and “myself,” an energy repository that offers a sense of security. The contradictions between people require an internal space and time for digestion. In the film, nature serves as a backdrop, which can be understood as a grand context reflecting social and political phenomena—a juxtaposition to my internal space. As for the dance movements, there are both improvisational and pre-choreographed elements. Body language serves not only as an expression but also as a presentation. It responds to what is happening in the current societal context that we cannot ignore. The body generates energy, both physically and visually. Without folding, my body forms a triangle from my head to my arms, and if you place a triangle within a circle, it will naturally appear sharper. What I do is create, present, and cultivate energy.

 

The New Benchmark for Efficient Language Model Deployment: Dynamic Hierarchical Transformer

The Chicago Journal Staff on October 25, 2023

A team from the Center for Deep Learning Northwestern University recently proposed a transformer architecture with adaptive online compression capabilities – Sample-based Dynamic Hierarchical Transformer (DHT). DHT is not only advantageous for adaptively optimizing the underlying network architecture during training but also has a flexible network for efficient inference. A unique aspect of their approach is the fact that the underlying layers and heads are sample-specific, exploiting the strategy that customizes the number of layers and heads for each single sample. Relevant results were published in “Proceedings on Engineering Sciences”. Northwestern University and the Institute of Computing Technology of the Chinese Academy of Sciences are the research and development units. The first author is PhD student Fanfei Meng. 

High efficiency, dynamic search, reinforcement learning-based tuning mechanism, and economical automatic machine learning paradigm have become hot topics recently. The transformer model, with self-attention layers and multi-head space, is considered a very promising deep language model due to its low computational complexity and excellent performance in many classification tasks. However, transformers still confront high computational costs and overfitting problems, which are challenged by serious computing consumption constraints in many business use cases. Therefore, how to design a model with high efficiency and low memory requirements is critical to conducting intensive computations on edges. According to reports, the team optimized the dynamic search mechanism via linear contextual bandits, showing that the training efficiency of Dynamic Hierarchical Transformer is generally increased by 74%, and the inference efficiency is increased by 81%. To the best of our knowledge, DHT is the most efficient Transformer model architecture in the world. 

Traditional network compression methods require full layer training first and then reduce the model size through layer-wise network compression or knowledge distillation. On the one hand, this two-step compression procedure can be of high time complexity. On the other hand, different tasks may require different lightweight transformers, making the uniform compression inflexible. By contrast, the team designed a dynamic, data-driven transformer model whose size can be optimized during training, skipping the separate compression step while maintaining a decent predicting capability. In addition, considering the effect of head interactions and the order samples appear during training, they formulate rewards of batch level rather than one-step gains, which successfully mitigates model performance reductions. 

Professor Archer, a faculty member of Cornell University, believes that the pioneering achievement has great potential in this field. This work provides a new idea for the industrial deployment of online compression and neural architecture search. The paradigm of sample-based adaptive training and inference down streaming owns excellent computing efficiency and lower hardware needs, which is promising for the development of edge computing and mobile computing.