Kir Belevich
deepsweet
AI & ML interests
MLX
Recent Activity
updated a dataset about 4 hours ago
deepsweet/MLX-oQ-sensitivity-map published a dataset about 4 hours ago
deepsweet/MLX-oQ-sensitivity-map updated a model about 20 hours ago
deepsweet/Qwen3.6-27B-MLX-VL-oQ4Organizations
None yet
Any chance to get deepsweet/Qwen3.6-27B-MLX-oQ6 :)
❤️ 1
3
#2 opened 7 days ago
by
evkt
Please upload a Qwen3.6-27B-MLX-VL-oQ8 for M3+
👀 1
1
#4 opened 10 days ago
by
adampotocki
404 error at https://huggingface.co/deepsweet/Qwen3.6-27B-MLX-VL-oQ6-FP16
❤️ 1
1
#1 opened 12 days ago
by
nreliu
Is there a benchmark comparison between DFLASH and native MTP?
3
#3 opened 13 days ago
by
crazyi
Please upload a Qwen3.6-27B-MLX-VL-oQ6 for M3
1
#2 opened 14 days ago
by
alexaione
Please build a Qwen3.6-35B-A3B-MLX-VL-oQ6
4
#1 opened 15 days ago
by
alexaione
There is no benchmark for mlx models
👍 1
7
#4 opened 22 days ago
by
zaskara
The original Qwen3.6-27B is multimodal
8
#1 opened 21 days ago
by
crazyi
Some questions
2
#2 opened 21 days ago
by
FaheemIqbal
Openclaw compatibility problem
1
#1 opened 21 days ago
by
felkf
I used it on Omlx,but it showed thinking as content.
3
#4 opened 23 days ago
by
BeCreated
non FP16 : M3+ version of VLM
2
#2 opened 25 days ago
by
jeromelaurens
M1 Ultra 64GB FP16 improvements
9
#1 opened 26 days ago
by
KnucklesXBT
Can we support image to text in this model?
3
#2 opened 28 days ago
by
mastervivi
🛠️ revert chat template to the original one
#1 opened about 1 month ago
by
deepsweet
🛠️ revert chat template to the original one
#1 opened about 1 month ago
by
deepsweet
🛠️ revert chat template to the original one
#1 opened about 1 month ago
by
deepsweet
Add LICENSE file
👍 1
#33 opened about 1 month ago
by
deepsweet
Thanks to you, M1 users cheered!
3
#1 opened about 1 month ago
by
felkf
What does the FP16 suffix signify?
3
#1 opened about 1 month ago
by
fisherwei