caoshiyi
Ph.D. student @ UC Berkeley. Previous MS @ ETHz, CS undergrad @ SJTU. AI & Sys
UC BerkeleyBerkeley, CA
Pinned Repositories
a3c_indigo
DRL in Network Congestion Control. Completion of the A3C implementation of Indigo based on the original Indigo codes. Tested on Pantheon.
FlexFlow
A distributed deep learning framework that supports flexible parallelization strategies.
FlexGen
Running large language models on a single GPU for throughput-oriented scenarios.
RSA_CCA2
Server Client model for Textbook RSA. CCA2 for Textbook RSA. OAEP RSA.
Simple_Dashboard
Flask + D3: A simple dashboard with Choropleth Map, Radar Chart and Bar Chart.
FlexFlow
FlexFlow Serve: Low-Latency, High-Performance LLM Serving
atlas
S-LoRA
S-LoRA: Serving Thousands of Concurrent LoRA Adapters
sglang
SGLang is a fast serving framework for large language models and vision language models.
vllm
A high-throughput and memory-efficient inference and serving engine for LLMs
caoshiyi's Repositories
caoshiyi/a3c_indigo
DRL in Network Congestion Control. Completion of the A3C implementation of Indigo based on the original Indigo codes. Tested on Pantheon.
caoshiyi/RSA_CCA2
Server Client model for Textbook RSA. CCA2 for Textbook RSA. OAEP RSA.
caoshiyi/Simple_Dashboard
Flask + D3: A simple dashboard with Choropleth Map, Radar Chart and Bar Chart.
caoshiyi/FlexFlow
A distributed deep learning framework that supports flexible parallelization strategies.
caoshiyi/FlexGen
Running large language models on a single GPU for throughput-oriented scenarios.
caoshiyi/ML-Self-Study
Machine Learning Self-study
caoshiyi/pspin
PsPIN: A RISC-V in-network accelerator for flexible high-performance low-power packet processing
caoshiyi/quartz
The Quartz Quantum Compiler
caoshiyi/AdaM
The source code for the ICPP'19 paper AdaM: An Adaptive Fine-Grained Scheme for Distributed Metadata Management.
caoshiyi/asciiclass
Notes and Labs for Advanced Topics in Data Processing
caoshiyi/caoshiyi.github.io
caoshiyi/databaseology
Collection of Papers On Database Management Systems
caoshiyi/vllm
A high-throughput and memory-efficient inference and serving engine for LLMs
caoshiyi/FastChat
An open platform for training, serving, and evaluating large language models. Release repo for Vicuna and Chatbot Arena.
caoshiyi/HyQuas
A hybrid partitioner based quantum circuit simulation system on GPU
caoshiyi/jax
Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more
caoshiyi/lm-sys.github.io
caoshiyi/SmartHome_WeChat
Smart Home System. Using WeChat as the client: make commands on or obtain environmental data from IoT deices.
caoshiyi/Traffic_Visualization
ChinaVis 2019. D3, PostgreSQL, Leaflet. Traffic Visualization.
caoshiyi/weeklog2018
This is the project of week logs of index groups in 2018.