Ecosyste.ms: Issues
An open API service for providing issue and pull request metadata for open source projects.
GitHub / axonn-ai/axonn issues and pull requests
#105 - add hatchet literal support for timers
Pull Request -
State: closed - Opened by S-Mahua 22 days ago
#104 - Add more timers
Pull Request -
State: closed - Opened by anishbh 23 days ago
- 1 comment
#103 - Disable unnecessary non expert comm
Pull Request -
State: closed - Opened by siddharth9820 about 1 month ago
#102 - Autotuning - initial infra + a small optimization for pure depth tp
Pull Request -
State: closed - Opened by siddharth9820 about 1 month ago
#101 - Adding timers with cct
Pull Request -
State: closed - Opened by siddharth9820 about 1 month ago
Labels: WIP
#100 - Minor changes for Release 0.2.0
Pull Request -
State: closed - Opened by siddharth9820 about 1 month ago
#99 - Correct url of ci tests badge
Pull Request -
State: closed - Opened by siddharth9820 about 1 month ago
#98 - Update .readthedocs.yml
Pull Request -
State: closed - Opened by siddharth9820 about 1 month ago
#97 - Allow Passing Model Keyword Arguments Through the parallelize Interface
Pull Request -
State: closed - Opened by Dando18 2 months ago
- 1 comment
Labels: enhancement, ready-for-review
#96 - reorg code and first implementation of the new easy API
Pull Request -
State: closed - Opened by siddharth9820 3 months ago
#95 - Improving AxoNN's memory consumption
Pull Request -
State: closed - Opened by siddharth9820 5 months ago
#94 - Tensor Parallel Embeddings
Issue -
State: open - Opened by siddharth9820 5 months ago
#93 - Tensor parallel embedding
Pull Request -
State: closed - Opened by siddharth9820 5 months ago
#92 - Tracking all issues related to litgpt
Issue -
State: open - Opened by siddharth9820 5 months ago
#91 - create an engine for all things pipelining and deprecate custom mixed precision
Pull Request -
State: closed - Opened by siddharth9820 5 months ago
#90 - Deprecate mixed precision support
Issue -
State: closed - Opened by siddharth9820 5 months ago
- 1 comment
#89 - Tracking all issues related to pytorch lightning support here
Issue -
State: open - Opened by siddharth9820 5 months ago
#88 - make no-grad-sync yield None
Pull Request -
State: closed - Opened by siddharth9820 5 months ago
- 1 comment
#87 - Improve CI tests
Issue -
State: open - Opened by siddharth9820 5 months ago
#86 - Create a finetuning and offline inference example in axonn-examples
Issue -
State: closed - Opened by siddharth9820 5 months ago
#85 - User Guide
Issue -
State: open - Opened by siddharth9820 5 months ago
#84 - Checkpoint Combiner
Issue -
State: open - Opened by siddharth9820 5 months ago
#83 - Supporting init_module, load/save checkpoint
Pull Request -
State: closed - Opened by siddharth9820 5 months ago
#82 - More lightning features
Pull Request -
State: closed - Opened by siddharth9820 5 months ago
#81 - Update advanced.rst
Pull Request -
State: closed - Opened by siddharth9820 5 months ago
#80 - User guide Changes
Pull Request -
State: closed - Opened by siddharth9820 5 months ago
#79 - Creating this PR to document all functions in the codebase
Pull Request -
State: open - Opened by siddharth9820 5 months ago
#78 - Changes to fix issues in IFT.
Pull Request -
State: closed - Opened by siddharth9820 6 months ago
#77 - Add API for tensor parallel model checkpointing
Pull Request -
State: closed - Opened by siddharth9820 6 months ago
#76 - AxonnStrategy for Lightning Fabric backend
Pull Request -
State: closed - Opened by anishbh 6 months ago
#75 - Merge develop into CPU branch
Pull Request -
State: closed - Opened by Avuxon 6 months ago
#74 - This PR shards the Dataloader across depth and data parallel ranks both
Pull Request -
State: closed - Opened by siddharth9820 7 months ago
#73 - initial doc for EasyAPI, Accelerate, and FT example
Pull Request -
State: closed - Opened by jwendlan 7 months ago
#72 - Make monkeypatching more efficient and change easy API to a single argument
Pull Request -
State: closed - Opened by siddharth9820 7 months ago
#71 - Convolution hot fix
Pull Request -
State: closed - Opened by siddharth9820 8 months ago
#70 - added automatic_parallelism
Pull Request -
State: closed - Opened by S-Mahua 8 months ago
- 1 comment
#69 - docs: fix build issues and add sub-sections
Pull Request -
State: closed - Opened by bhatele 9 months ago
#68 - Bugfix: Initialize grad_input, grad_weight to None
Pull Request -
State: closed - Opened by adityaranjan 9 months ago
Labels: ready-for-review
#67 - change parallelize context to use AutoConfig
Pull Request -
State: closed - Opened by siddharth9820 9 months ago
#66 - Removing the drop and gathers in depth tensor parallelism for the easy API
Pull Request -
State: closed - Opened by siddharth9820 9 months ago
#65 - adding parallelize context for opt
Pull Request -
State: closed - Opened by jwendlan 9 months ago
#64 - Add communication dtypes for all-gathers and reduce scatters in depth tensor parallelism
Pull Request -
State: open - Opened by siddharth9820 9 months ago
#63 - removed mpi4py dependency
Pull Request -
State: closed - Opened by S-Mahua 9 months ago
#62 - Change overlap for depth tp and do not initialize MPI unless absolutely needed
Pull Request -
State: closed - Opened by siddharth9820 10 months ago
#61 - Parallel Transformers
Pull Request -
State: closed - Opened by jwendlan 11 months ago
- 1 comment
#60 - Added Depth Tensor Parallelism to Conv Layer
Pull Request -
State: closed - Opened by prajwal1210 11 months ago
Labels: ready-for-review
#59 - Parallel transformers
Pull Request -
State: closed - Opened by jwendlan 11 months ago
#58 - storing parallel hf implementations in axonn
Pull Request -
State: closed - Opened by jwendlan 11 months ago
#57 - More communication optimizations
Pull Request -
State: closed - Opened by siddharth9820 11 months ago
#56 - Rebase axonn-cpu to master
Pull Request -
State: closed - Opened by Avuxon 12 months ago
#55 - Fixing some issues with depth tensor parallelism
Pull Request -
State: closed - Opened by siddharth9820 12 months ago
- 2 comments
#54 - A context manager to optimize communication
Pull Request -
State: closed - Opened by siddharth9820 about 1 year ago
- 2 comments
#53 - change outer variables
Pull Request -
State: closed - Opened by siddharth9820 about 1 year ago
#52 - add option to change batch dimension in drop
Pull Request -
State: closed - Opened by siddharth9820 about 1 year ago
#51 - Initialize layers on the GPU
Pull Request -
State: closed - Opened by siddharth9820 about 1 year ago
#50 - Make mpi4py an optional dependency
Issue -
State: closed - Opened by siddharth9820 about 1 year ago
#49 - first iteration of 3D tensor parallelism
Pull Request -
State: closed - Opened by siddharth9820 about 1 year ago
- 1 comment
#48 - Visualize topology of GPUs
Issue -
State: open - Opened by siddharth9820 about 1 year ago
#47 - Repair broke CI test logo
Issue -
State: open - Opened by siddharth9820 about 1 year ago
#46 - Adding third dimension of intra-layer parallelism
Pull Request -
State: closed - Opened by siddharth9820 about 1 year ago
#45 - Add Easy API to convolution layers
Issue -
State: open - Opened by siddharth9820 about 1 year ago
#44 - Intra-layer - Overlap communication in backward pass
Pull Request -
State: closed - Opened by siddharth9820 about 1 year ago
- 3 comments
#43 - Test AxoNN with Pytorch 2.0+
Issue -
State: closed - Opened by siddharth9820 about 1 year ago
- 1 comment
#42 - Make pipeline parallelism modular
Issue -
State: closed - Opened by siddharth9820 about 1 year ago
#41 - add dependencies between workflows
Pull Request -
State: closed - Opened by bhatele about 1 year ago
- 1 comment
Labels: ready-for-review
#40 - [WIP] A tensor parallel API for beginners
Pull Request -
State: closed - Opened by siddharth9820 about 1 year ago
- 2 comments
Labels: WIP
#39 - Adding CPU training support to AxoNN
Pull Request -
State: open - Opened by Avuxon about 1 year ago
- 2 comments
Labels: ready-for-review
#38 - [WIP] ILP Conv Layer support
Pull Request -
State: closed - Opened by prajwal1210 about 1 year ago
- 2 comments
Labels: WIP
#37 - [WIP] add bfloat16 training support
Pull Request -
State: closed - Opened by siddharth9820 about 1 year ago
- 1 comment
Labels: ready-for-review
#36 - changes to the intra-layer API for the GPT benchmark
Pull Request -
State: closed - Opened by siddharth9820 about 1 year ago
#35 - Only initialize inter-layer if G_inter > 1
Pull Request -
State: closed - Opened by siddharth9820 about 1 year ago
- 1 comment
#34 - add AxoNN logo
Pull Request -
State: closed - Opened by bhatele about 1 year ago
#33 - CI/CD tests for intra-layer parallelism
Pull Request -
State: closed - Opened by siddharth9820 over 1 year ago
#32 - add autocasting capabilities
Issue -
State: closed - Opened by siddharth9820 over 1 year ago
#31 - readme: add slack link
Pull Request -
State: closed - Opened by bhatele over 1 year ago
#30 - add 2D tensor parallelism for FC layers
Pull Request -
State: closed - Opened by siddharth9820 over 1 year ago
#29 - Docs: installation and running mnist test
Pull Request -
State: closed - Opened by adityaranjan almost 2 years ago
#28 - Tests: convert memopt to int before bool
Pull Request -
State: closed - Opened by adityaranjan almost 2 years ago
#27 - fix g_intra print
Pull Request -
State: closed - Opened by zsat about 2 years ago
#26 - docs: fix readthedocs.org build issues
Pull Request -
State: closed - Opened by bhatele about 2 years ago
#25 - Add wall clock breakdown
Pull Request -
State: closed - Opened by siddharth9820 about 2 years ago
#24 - add checkpointing and post backward hook support
Pull Request -
State: closed - Opened by siddharth9820 over 2 years ago
#23 - No need to have two gradient buffers for mixed precision
Issue -
State: closed - Opened by siddharth9820 over 2 years ago
#22 - support gradient clipping
Issue -
State: closed - Opened by siddharth9820 over 2 years ago
#21 - Support for intra-layer parallelism
Pull Request -
State: closed - Opened by siddharth9820 over 2 years ago
#20 - init should also set the microbatch size
Issue -
State: closed - Opened by siddharth9820 over 2 years ago
#19 - Coalesce and reassign is memory inefficient
Issue -
State: closed - Opened by siddharth9820 over 2 years ago
#18 - fix evaluation bug for inter-layer
Pull Request -
State: closed - Opened by siddharth9820 over 2 years ago
#17 - Bug in full precision
Issue -
State: closed - Opened by siddharth9820 over 2 years ago
#16 - Update README
Pull Request -
State: closed - Opened by bhatele over 2 years ago
#15 - add structure for AxoNN docs
Pull Request -
State: closed - Opened by bhatele over 2 years ago
#14 - Release 0.1.0
Pull Request -
State: closed - Opened by siddharth9820 over 2 years ago
#13 - Release 0.0.1
Pull Request -
State: closed - Opened by siddharth9820 over 2 years ago
- 2 comments
#12 - Reproduce IPDPS results
Pull Request -
State: closed - Opened by siddharth9820 almost 3 years ago
#11 - Large LM training
Pull Request -
State: closed - Opened by siddharth9820 almost 3 years ago
#10 - Add validation/testing support
Pull Request -
State: closed - Opened by siddharth9820 almost 3 years ago
#9 - Implementation of the cpu-offloading memory optimizations
Pull Request -
State: closed - Opened by siddharth9820 almost 3 years ago
#8 - Corrected bug in _sync_scale, changed print_status
Pull Request -
State: closed - Opened by siddharth9820 almost 3 years ago
#7 - AxoNN's implementation of mixed precision for hybrid parallelism
Pull Request -
State: closed - Opened by siddharth9820 almost 3 years ago
#6 - Adding mixed precision support
Pull Request -
State: closed - Opened by siddharth9820 almost 3 years ago
Labels: WIP