Hi there i'm aaron.
Hi there, hope you enjoy your stay. I'm a software engineer who is curious about system design, and tools around deep neural nets. Occasionally I do climb some rocks.
aarnphm
Follow
Pinned
2,950 contributions in the last year
Activity overview
Contribution activity
June 2023
Created 254 commits in 5 repositories
Created a pull request in hwchase17/langchain that received 7 comments
feat: OpenLLM
+1,196
−87
•
7
comments
Opened 23 other pull requests in 3 repositories
bentoml/BentoML
2
open
7
merged
2
closed
- fix(json): only show warning if custom json encoder is used
- docs: update wrong instruction for tracing
- feat(scheduling): support assigning multiple GPU per worker instance
- fix(strategy): handle special case when assigning device 0
- chore(ci): disable windows e2e
- fix(transformers): Lazily check pretrained type for attribute
- fix(server): remove setting working_dir to None
- feat(cli): fix -o tag to quiet all logs
- chore(cli): add spaces between aliases
- fix: lazy load types
- fix(configuration): patch keysize changing during iteration
bentoml/OpenLLM
8
merged
1
closed
1
open
huggingface/transformers
2
open
Reviewed 44 pull requests in 4 repositories
bentoml/BentoML
25 pull requests
- feat(bentocloud): implement deployment CLI
- docs: update wrong instruction for tracing
- fix: remove legacy metrics
- chore(deps): bump pyarrow from 12.0.0 to 12.0.1
- fix(ci): the timeout issue on Windows e2e and re-enable
- chore: use click command groups for bento subcommand
- fix: correctly handle 404 in Yatai REST API client
- chore: remove debug lines
- fix(serialization): loads gpu torch tensor on cpu-only machine and revert DefaultContainer optimization
- improv(framework): diffusers sync BentoModel version with huggingface version
- feat: support ipv6 addresses for serve
- feat(bentocloud): implement update, terminate and delete deployment
- chore(deps): bump transformers from 4.11.3 to 4.30.0 in /examples/inference_graph
- chore(deps): bump bufbuild/buf-setup-action from 1.20.0 to 1.21.0
- chore(deps): bump pytest from 7.3.1 to 7.3.2
- docs: Fix wording in intro tutorial
- doc: Fix the broken link to og image
- feat(scheduling): support assigning multiple GPU per worker instance
- fix(strategy): handle special case when assigning device 0
- feat: Service lifecycle hooks
- bugs: Fix SKLearn pipeline example
- feat: cleaner APIs for testing Service implementation
- feat: Support Create deployment
- fix(server): remove setting working_dir to None
- fix: read uncompressed bentos
- Some pull request reviews not shown.
bentoml/OpenLLM
4 pull requests
ayamir/nvimdots
2 pull requests
hwchase17/langchain
1 pull request
Created an issue in bentoml/OpenLLM that received 1 comment
perf: openllm.LLM metaclass creation
Feature request Currently, the metaclass implementation are rudimentary and inefficient. It actually takes quite a bit to initialize the model, giv…
1
comment
Opened 8 other issues in 4 repositories
bentoml/OpenLLM
1
open
3
closed
bentoml/BentoML
2
open
hwchase17/langchain
1
open
huggingface/transformers
1
open
Started 1 discussion in 1 repository
aarnphm/whispercpp
aarnphm/whispercpp
-
General updates and looking for usage!
This contribution was made on Jun 4
1
contribution
in private repositories
Jun 10