1.
welcome
2.
chronicles
❱
2.1.
feb 2024
3.
projects
❱
3.1.
copilot
3.2.
library
4.
survey
❱
4.1.
pollux
4.2.
adasum
4.3.
adaptation_learning
4.4.
gradient_descent
4.5.
auto_parallel
4.6.
scheduling
4.7.
gradient_compression
❱
4.7.1.
dgc
4.7.2.
csc
4.8.
flash attention
4.9.
LoRA
5.
mathematics
❱
5.1.
basic
5.2.
entropy
5.3.
newton
5.4.
regression
5.5.
conjugate descent
5.6.
gradient descent
5.7.
pca
5.8.
support vector
5.9.
differentiation
5.10.
fourier
5.11.
kmeans
6.
wavelets
❱
6.1.
plan
6.2.
preliminary
6.3.
haar wavelet
6.4.
fourier analysis
6.5.
uncertainty principle
6.6.
multiresolution
7.
models
❱
7.1.
llm
7.2.
falcon
7.3.
llama
7.4.
peft
7.5.
transformer
7.6.
models
8.
megatron
9.
deepspeed
10.
pytorch
❱
10.1.
tensor
10.2.
autograd
10.3.
operator
10.4.
profiler
10.5.
hook
10.6.
elastic
10.7.
patch
10.8.
misc
11.
paddle
❱
11.1.
ps
11.2.
framework
11.3.
cinn
11.4.
dataloader
12.
horovod
❱
12.1.
run
12.2.
workflow
12.3.
object
12.4.
develop
12.5.
pytorch
12.6.
tensorflow
12.7.
elastic
13.
ray
❱
13.1.
overview
13.2.
gcs
13.3.
raylet
13.4.
api
13.5.
survey
14.
python
❱
14.1.
concurrent execution
14.2.
multiprocessing
14.3.
decorator
15.
tips
❱
15.1.
enable_shared_from_this
15.2.
ip_local_port_range
15.3.
golang error
16.
infrastructure
❱
16.1.
pki
16.2.
linux cache
17.
kubernetes
❱
17.1.
concepts
17.2.
scheduler
17.3.
operator
17.4.
device plugin
17.5.
docker
17.6.
install
17.7.
api-service
18.
nccl
19.
cuda
20.
todo
❱
20.1.
gloo
20.2.
mpi
20.3.
jax
20.4.
tvm
20.5.
llm
21.
notes
❱
21.1.
influence and persuasion
21.2.
freynman technique
21.3.
wavelet signal processing
Light
Rust
Coal
Navy
Ayu
Aller au boulot
copilot
Goals
Run local copilot with llm.
References
https://developer.apple.com/metal/pytorch/
https://github.com/vllm-project/vllm