Eugene Klimov
Slach
AI & ML interests
None yet
Recent Activity
liked a model 3 days ago
ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF upvoted a paper 3 days ago
Disaggregated Quantization: Specializing LLM Prefill and Decode liked a model 8 days ago
TypeSafeAI/Step-5-Preview-BF16Organizations
performance still the same on DGX Spark and latest pre-release llama.cpp on DGX Spark GB10
#6 opened about 1 month ago
by Slach
Q4_K_M wrong number of tensors; expected 81, got 58
➕ 5
4
#4 opened about 1 month ago
by 
JMorris1999
mtp.* tensors look like random init, not trained weights
🧠👍 9
17
#10 opened about 1 month ago
by 
EryriLabs
New activity in georgeis55/Nemotron-3.5-Lightning-30B-A3B-refusal-ablated-MLX-oQ8-mtp about 2 months ago
try to use it with omlx (0.5.8+) got error
#1 opened about 2 months ago
by Slach
tried this model with omlx and tools never calls
1
#1 opened about 2 months ago
by Slach
github link not found
1
#1 opened about 2 months ago
by Slach
if i have russian or armenian audio dataset, how to fine tune this model?
#1 opened 3 months ago
by Slach
MTP weights
➕❤️ 26
4
#4 opened 3 months ago
by 
scottgl
can't load model in lm-studio.ai MLX latest version
3
#1 opened 3 months ago
by Slach
doesn't work in `lmsysorg/sglang:dev-cu13` on nvidia dgx spark
👍 2
1
#1 opened 7 months ago
by Slach
is this model still can use 1M context window or not?
1
#1 opened 9 months ago
by Slach
any plan about GLM-4.7-REAP-139B-FP8?
3
#3 opened 9 months ago
by Slach
Could we create REAP version of this models with MXFP4 ?
1
#2 opened 9 months ago
by Slach
could we create 139B quantization with REAP for this model?
5
#2 opened 9 months ago
by Slach