Allen Nie

Not able to run example file

2

Hi, It seems that you might have written this over a year ago, so sorry to bother you...I'm running with TF 0.12.1, which is released a year ago (hopefully the...

Batched Computation

4

Hi Ryan, The LSH based forward and backward computation is a great idea and your paper has a nice contribution! However, you have to sample from the LSH buckets (different...

和Stanford Word Segmenter比较，以及如何在论文中引用？

1

打扰，请问这个工具用的算法和Stanford Word Segmenter用的算法有什么主要区别吗？ 以及，如果使用了这个工具，需要引用你们的论文吗？（或者说这个工具有基于某篇论文吗？）——我也可以直接引用github链接。。。

Why initialize W[0] with all 0s?

4

Hi, I'm just having some trouble understanding the process_data.py file, especially saving a special W[0] word and initialize idx_map starting at 1. What's the purpose of doing that?

Even though `model.load("checkpoint.pkl")` can already produce a runnable model, sometimes, there is a need to submit a runnable Python agent to an external evaluation service (such as various competitions that...

Add reset() to trace.model

4

Multiple people asked me if there's a simple way to reset parameter node value back to original. Will work on adding this.

LLM backend signature not unified

3

Currently LLM backend `__init__` signature is not unified (this is a legacy backward compatibility issue mostly with AutoGen). For three backends, the signatures are: ```python class AutoGenLLM(AbstractModel): def __init__(self, config_list:...