Ton Cao PRO
AI & ML interests
Democratizing LLM @cyankiwi
Recent Activity
updated a model about 3 hours ago
cyankiwi/Swift-Qwen3.8-27B-AWQ-BF16-INT4 updated a model 1 day ago
cyankiwi/Swift-Qwen3.8-27B-AWQ-INT4 updated a model 2 days ago
cyankiwi/MiMo-V2.6-Distill-Qwen-9B-AWQ-BF16-INT4Organizations
Fix MTP ignore names for SGLang fused Linear layers (`qkv_proj` / `gate_up_proj`)
1
#9 opened 16 days ago
by
dived42
Loading issue on ROCm — asymmetric `compressed-tensors` MoE not supported in vLLM
1
#2 opened 18 days ago
by
tensornet
Sync chat_template.jinja with upstream google/gemma-4-26B-A4B-it
1
#10 opened 20 days ago
by
Johnson145
layer 44's 288 experts are left bf16
2
#1 opened 29 days ago
by
MirecX
Running Qwen3.8-Flash-Next-AWQ-INT4
12
#1 opened 30 days ago
by
cpatonn
Do AWQ-BF16-INT4
➕ 9
7
#1 opened about 1 month ago
by
ab22k
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
1
#3 opened about 1 month ago
by
zhnagchenchne
Tool Calling and Reasoning Question
👀 1
2
#1 opened about 2 months ago
by
kalashshah19
Looking forward to testing this, thanks!
9
#1 opened 3 months ago
by
dnhkng
Loops
2
#1 opened 2 months ago
by
valentijnvenus
AWQ Dataset
2
#1 opened 3 months ago
by
Dsturb
generation settings???
2
#6 opened 3 months ago
by
yuchenxie
Wrong files?
1
#1 opened 3 months ago
by
ztsvvstz
F16 or BF16?
6
#6 opened 4 months ago
by
qenme
Is there a plan to upload the M3 model?
👍 1
1
#4 opened 3 months ago
by
chengongliang
how to use it
10
#1 opened 4 months ago
by
iSource1
error trying to run mini on a single 5090
1
#1 opened 4 months ago
by
robert896r1
Unable to serve model, got "Only symmetric quantization is supported for MoE" using vllm 0.22.1
1
#1 opened 4 months ago
by
ricky-cck
Vllm and SgLang command please
👍 1
5
#1 opened 4 months ago
by
mtcl
New Safetensors upload
2
#8 opened 4 months ago
by
meganoob1337