idiap/HMMGradients.jl

Enables computing the gradient of the parameters of Hidden Markov Models (HMMs)

Juliajuliaspeech-recognitiongradienthidden-markov-modelsctc-losshmmshmm-training
This is stars and forks stats for /idiap/HMMGradients.jl repository. As of 10 May, 2024 this repository has 22 stars and 0 forks.

HMMGradients.jl This package enables computing the gradient of the parameters of Hidden Markov Models (HMMs). This makes it possible to perform HMM training using gradient based methods like stochastic gradient descent, which is necessary for example when neural networks are involved, e.g. in modern automatic speech recognition systems. Check out this TIDIGITS recipe for an example. Formally, this package extends ChainRulesCore making it possible to train HMM models using the automatic differentiation...
Read on GithubGithub Stats Page
repotechsstarsweeklyforksweekly
Suzhou-Tongyuan/jnumpyJuliaPythonShell224080
SciFracX/FractionalDiffEq.jlJulia67070
CliMA/ClimaTimeSteppers.jlJulia42+130
Merck/GeneratedExpressions.jlJulia8010
DeutscheKI/tevr-asr-toolCC++Other3990170
leon-ai/leonTypeScriptJavaScriptPython13.6k01.1k0
ChenNingCong/BuildSystemJuliaShell17010
Drvi/ProtocolBuffers.jlJulia4000
invenia/UTCDateTimes.jlJulia7000
jbytecode/Sqlite3Stats.jlJulia17030