issues
search
redhat-et
/
ilab-on-ocp
12
stars
16
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Demo img to main branch
#158
cooktheryan
closed
4 hours ago
0
Mt_bench step uses 3 gpu
#157
cooktheryan
closed
57 minutes ago
0
add simple chatbot definition
#156
sallyom
opened
1 day ago
1
feat: pass data between tasks via PVCs
#155
tumido
closed
1 hour ago
0
reset sampling size for training
#154
sallyom
opened
1 day ago
1
Tune pcd sampling size
#153
MichaelClifford
closed
1 day ago
0
switch to the granite provided by rhel ai rather than huggingface
#152
cooktheryan
opened
1 day ago
0
update Containerfile
#151
sallyom
closed
1 day ago
0
WIP: hugging face to pvc
#150
cooktheryan
closed
1 day ago
0
chore: fix comment typo
#149
HumairAK
closed
4 days ago
0
chore: add default args for sdg pipeline
#148
HumairAK
closed
5 days ago
0
use separate secrets & configmaps and/or values for judge & teacher model servers
#147
sallyom
opened
5 days ago
0
add pipeline parameter for sdg
#146
sallyom
closed
5 days ago
0
Add pipeline parameter for SDG
#145
MichaelClifford
opened
5 days ago
0
increase kubectl_wait_for_op timeout
#144
sallyom
closed
5 days ago
0
backport #142: fix: use sub tmp dir
#143
leseb
closed
6 days ago
0
fix: use sub tmp dir
#142
leseb
closed
6 days ago
1
Backport #140
#141
leseb
closed
6 days ago
0
standalone: small fixes
#140
leseb
closed
6 days ago
0
remove namespace and further information
#139
cooktheryan
closed
6 days ago
0
Add variable for caching pipeline steps
#138
MichaelClifford
opened
1 week ago
4
Clean up commented out MMLU code in pipeline.yaml
#137
MichaelClifford
opened
1 week ago
0
Add a "cache" parameter to the pipeline to easily toggle this feature in KFP between runs
#136
MichaelClifford
opened
1 week ago
2
feat: allow using custom certificates for serving (backport #128)
#135
leseb
closed
1 week ago
1
fix: use separate model checkpoint directories for each training phase
#134
MichaelClifford
closed
1 week ago
2
chore: add pipeline-type config for sdg in standalone
#133
HumairAK
closed
5 days ago
1
serving with a pvc and directions
#132
cooktheryan
closed
1 week ago
0
Make sure each phase of training stores checkpoints in appropriate directories in KFP
#131
MichaelClifford
closed
1 week ago
0
add env vars for permissions
#130
sallyom
closed
1 week ago
0
feat: use venv feature from KFP
#129
leseb
opened
1 week ago
5
feat: allow using custom certificates for serving
#128
leseb
closed
1 week ago
2
update pipeline image
#127
sallyom
closed
1 day ago
3
fixes for pipeline - batch_size
#126
sallyom
closed
5 days ago
1
Rename pipeline parameters for better organization on the DSP run page
#125
MichaelClifford
opened
1 week ago
0
Update Repo README
#124
MichaelClifford
opened
1 week ago
0
add judge model details to standalone readme
#123
sallyom
opened
2 weeks ago
0
Add training args to pipeline
#122
MichaelClifford
closed
1 week ago
0
fix: improve pytorchjob in kfp
#121
MichaelClifford
closed
2 weeks ago
0
feat: implement SDG in cluster for standalone mode
#120
leseb
closed
6 days ago
6
docs: more explicit serving endpoint requirement
#119
leseb
closed
2 weeks ago
0
doc: add e2e flow diagram for standalone
#118
leseb
closed
2 weeks ago
0
fix: honour judge secret name
#117
leseb
closed
2 weeks ago
0
small clean up for mt_bench
#116
MichaelClifford
closed
2 weeks ago
0
Add check for existing `candidate_model`
#115
MichaelClifford
closed
2 weeks ago
0
Ensure evaluation steps skip *.jsonl files
#114
MichaelClifford
closed
2 weeks ago
1
Use knowledge data for phase 1 training and skills data for phase 2 training in standalone script
#113
MichaelClifford
closed
2 weeks ago
0
fix: do not use kfp-model-server anymore
#112
leseb
closed
2 weeks ago
0
fix: small typo
#111
leseb
closed
2 weeks ago
0
standalone: make change_dsl_function_to_normal_function() more robust
#110
leseb
opened
2 weeks ago
0
SDG on standalone
#109
leseb
closed
6 days ago
1
Next