Directly distill from Llama, the finetune in DPO
Junxiong Wang
JunxiongWang
AI & ML interests
Attention Free Model / Subquadratic Language Models
Recent Activity
updated
a model
2 days ago
togethercomputer/MambaInLlama3B
updated
a model
3 days ago
togethercomputer/MambaInLlama1B
updated
a dataset
7 days ago
JunxiongWang/model_revision_max_4_closest_and_random
Organizations
Collections
7
models
37
JunxiongWang/mamba_0_5_distill
Updated
•
10
JunxiongWang/Llama3.2-Mamba-3B-dpo
Updated
•
5
JunxiongWang/Llama3.2-Mamba-3B-distill
Updated
•
33
JunxiongWang/Llama3.2-Mamba2-3B-distill
Updated
•
91
JunxiongWang/Llama3.2-Mamba2-3B-dpo
Updated
•
13
JunxiongWang/Llama3.1-Mamba2-8B-dpo
Updated
•
1
JunxiongWang/Llama3.1-Mamba-8B-dpo
Updated
JunxiongWang/Llama3.1-Mamba2-8B-distill
Updated
JunxiongWang/Llama3.1-Mamba-8B-distill
Updated
JunxiongWang/MambaByte_Stories
Text Generation
•
Updated
•
4
•
1
datasets
5
JunxiongWang/model_revision_max_4_closest_and_random
Viewer
•
Updated
•
530k
•
39
JunxiongWang/sftdatasetv3
Viewer
•
Updated
•
12.4M
•
535
JunxiongWang/sftdataset
Viewer
•
Updated
•
11M
•
96
•
2
JunxiongWang/llama3-ultrafeedback-armorm
Viewer
•
Updated
•
61.8k
•
79
•
1
JunxiongWang/testdataset
Viewer
•
Updated
•
1M
•
57