Raushan Turganbay
RaushanTurganbay
AI & ML interests
Generation and Multimodality
Recent Activity
liked a model about 19 hours ago
microsoft/Mage-VL liked a model about 19 hours ago
OpenMOSS-Team/MOSS-VL-Realtime liked a model 1 day ago
moonshotai/Kimi-K3Organizations
Update tokenizer_config.json
#14 opened 12 days ago
by
RaushanTurganbay
Fix config model type
#22 opened 12 days ago
by
RaushanTurganbay
Optimize group beam search by sharing the initial prefill across beams
3
#6 opened 23 days ago
by
lavrenko
"sink_cache" fails on Transformers 5.12 due to outdated "Cache.__init__" API
1
#1 opened 29 days ago
by
lavrenko
constrained-beam-search produces degenerate output with use_cache=True
3
#2 opened about 1 month ago
by
lavrenko
DoLa custom generation produces gibberish with use_cache=True
1
#1 opened about 1 month ago
by
lavrenko
Fix Transformers v5 cached decoding in DoLa
2
#2 opened 30 days ago
by
lavrenko
Support Transformers v5 cache handling
2
#3 opened 30 days ago
by
lavrenko
Custom group-beam-search changes deterministic output when cache is enabled
3
#3 opened about 1 month ago
by
lavrenko
Support Transformers v5 cache handling
2
#4 opened about 1 month ago
by
lavrenko
[Question] Why does LLaVA evaluation assert batch_size == 1 for benchmark inference?
1
#62 opened 2 months ago
by
austinkomachi
Compatibility with v5
#23 opened 4 months ago
by
RaushanTurganbay
Compatibility with v5
#10 opened 4 months ago
by
RaushanTurganbay
Compatibility with v5
#7 opened 4 months ago
by
RaushanTurganbay
Compatibility with v5
#5 opened 4 months ago
by
RaushanTurganbay
Compatibility with v5
#6 opened 4 months ago
by
RaushanTurganbay
Compatibility with v5
#6 opened 4 months ago
by
RaushanTurganbay
Compatibility with v5
#4 opened 4 months ago
by
RaushanTurganbay
Compatibility with v5
#4 opened 4 months ago
by
RaushanTurganbay