Surya Bhupatiraju
suryabhupa
AI & ML interests
part of the Gemma Team -- Language models, Reinforcement Learning
Organizations
suryabhupa's activity
inquiry for gemma-7b : d_model
1
#61 opened 10 months ago
by
seongwoon
Hallucinations, misspellings etc. Something seems broken?
21
#10 opened 6 months ago
by
sam-paech
tokenizer chat_template has no role system
2
#9 opened 6 months ago
by
wnma3mz
Citation URL redirects to Gemma-1
1
#8 opened 6 months ago
by
yumemio
Asking same thing twice or thrice in hugging face chat breaks it , same thing on ollama
1
#7 opened 6 months ago
by
Jayakumark
transformers load fails?
7
#6 opened 6 months ago
by
bdambrosio
Flash attention 2 is not working
3
#9 opened 6 months ago
by
nalf3in2
Unable to reproduce the score of gemma_2b at pass@1 in humaneval.
3
#53 opened 8 months ago
by
ChiYuqi
What do they mean by maj@1 ?
3
#44 opened 7 months ago
by
joserass
Fine-Tune a gemma model for question answering
17
#62 opened 10 months ago
by
Iamexperimenting
save, loading and inferencing the Gemma model
13
#64 opened 10 months ago
by
Iamexperimenting
Need info on pre-training and instruction-tuning data
3
#64 opened 10 months ago
by
markding
Inference with RTX 3090 got OOM
3
#89 opened 8 months ago
by
kathylee
Weird Performance Issue with Gemma-7b compared to Gemma-2b with Qlora
6
#91 opened 8 months ago
by
UserDAN
What's the context window for this model?
6
#73 opened 10 months ago
by
siddheshgunjal
pretraining Gemma for domain dataset
8
#41 opened 9 months ago
by
Iamexperimenting
gemma -2b with multi-gpu
3
#44 opened 9 months ago
by
Iamexperimenting
<pad> spam issue
13
#40 opened 10 months ago
by
Zewsic
evaluation loss not calculated during during?
2
#43 opened 9 months ago
by
Iamexperimenting
Dont download, google scuttled this model
16
#77 opened 9 months ago
by
Tom-Neverwinter