Samuel Azran
SamuelAzran
AI & ML interests
None yet
Recent Activity
liked
a dataset
1 day ago
SLPRL-HUJI/HebDB
liked
a dataset
14 days ago
fixie-ai/common_voice_17_0_timestamps
liked
a dataset
14 days ago
kika2000/synthetic-speaker-diarization-dataset2
Organizations
None yet
SamuelAzran's activity
thanks , how to fine tune?
19
#1 opened about 1 month ago
by
NickyNicky

New Gemma 2 27B?
2
#3 opened 10 months ago
by
SamuelAzran
Was it train after the latest Huggingface Transformers Gemma fix? if not any update plans?
#4 opened about 1 year ago
by
SamuelAzran
Should not be called mixtral, the models made into the moe are yi based
9
#2 opened over 1 year ago
by
teknium

How does the MoE work?
3
#5 opened over 1 year ago
by
PacmanIncarnate
One or two models during inference?
3
#3 opened over 1 year ago
by
Venkman42

You know Mixtral, Llama 2 70b, GPT3.5... Are All Much Better
1
#13 opened over 1 year ago
by
deleted
Awesome- Could you help with pointers on doing same for Other languages(Swedish)?
3
#2 opened over 1 year ago
by
Olofp
QLora or full fine-tuning?
1
#1 opened over 1 year ago
by
SamuelAzran
Was system message used during training?
1
#8 opened over 1 year ago
by
SamuelAzran
NEW! OpenLLMLeaderboard 2023 fall update
20
#356 opened over 1 year ago
by
clefourrier

Did you do full model fine tuning (all layers) or only adapters?
1
#2 opened over 1 year ago
by
SamuelAzran
Can you release a chat version soon ?
11
#8 opened over 1 year ago
by
dong0213
Great work, but why only 2048 context length?
1
#4 opened over 1 year ago
by
SamuelAzran
Would it work well with sequence length > 2048?
2
#1 opened almost 2 years ago
by
SamuelAzran
Thank you very much!
10
#2 opened almost 2 years ago
by
AiCreatornator
Error running the example code
21
#6 opened about 2 years ago
by
will33am
