Blog Β· corpus.blog/blogs/solmaz.io/posts
solmaz.io
solmaz.io
2026
13 Sept 2026
13 Sept 2026
DeepSeek V4.1 Flash looks contaminated with public benchmarks. You can see it in the data they...original β
13 Sept 2026
A threshold was crossed with DeepSeek V4.1 Flash, similar to when Claude Code launched, or the...original β
12 Sept 2026
Another deepseek v4.1 flash issue, model started to refuse work around 700k token line, saying...original β
12 Sept 2026
11 Sept 2026
Not sure if this is chain of thought degradation or regular, but I'm getting this no deepseek...original β
11 Sept 2026
10 Sept 2026
10 Sept 2026
"Fair challenge", "Fair β that's on me", "That error is the whole bug", "Short answer: no. Not...original β
10 Sept 2026
DeepSeek-V4.1-Flash is available on Hugging Face inference providers through @novita_labs on...original β
10 Sept 2026
10 Sept 2026
Running into out-of-memory freeze issues and having to hard-reset your DGX Spark while trying...original β
10 Sept 2026
10 Sept 2026
If you've missed out (like me), a new tiny model MiniCPM5 2B is topping its category, seems to...original β
9 Sept 2026
7 Sept 2026
Struggling to create the level of quality of 3D modeling some of the OAI employees and early...original β
7 Sept 2026
6 Sept 2026
6 Sept 2026
4 Sept 2026
3 Sept 2026
The vibeslopped UI will improve, but I can now say ourmodels.cc has finally caught up with...original β
2 Sept 2026
I see pi-workflows viewers growing, so wanted to give a disclaimer: I did major refactors and...original β
2 Sept 2026
1 Sept 2026
I went spelunking in some old hard drives and found some of my writing that goes 20 years back...original β
31 Aug 2026
I'm training a small model over my tweets and other writing and it's going....interestingoriginal β
31 Aug 2026
well we wanted a helpful, harmless and HONEST assistant, I don't know why yall are surprisedoriginal β
28 Aug 2026
28 Aug 2026
the year was 1999. Star Wars DroidWorks let you built custom robots and use them on different...original β
28 Aug 2026
27 Aug 2026
glm 5.3 flash at an antirez-style asymmetric q2 quant could plausibly fit on a single dgx spark...original β
27 Aug 2026
Qwen 3.8 Flash-Next will have to be quantized down to 4 bit precision or lower in order to fit...original β
26 Aug 2026
People want to migrate from CLAUDE md so much, a blog post I wrote on it last year is still one...original β
25 Aug 2026
Pi folk, there is an issue with compaction in pi that might drain your codex limits fasteroriginal β
25 Aug 2026
25 Aug 2026
This actually makes the most sense to me from what I've read till now. Wouldn't be surprised if...original β
22 Aug 2026
my @pidotdev extensions give me so much joy. so much tedium that I automated away, that I know...original β
22 Aug 2026
22 Aug 2026
This is your daily reminder that you can just vibe your dream terminal dashboards with...original β
20 Aug 2026
19 Aug 2026
Before someone says it, I know github.com/earendil-worksβ¦ exists. But it doesn't give me a CLI...original β
18 Aug 2026
It seems my tweet ragebaited some unintentionally (best kind of ragebait, with plausible...original β
18 Aug 2026
18 Aug 2026
Qwen 3.8 27b weights are probably the most civilization-changing 20gb of data published, until...original β
18 Aug 2026
18 Aug 2026
.@odd_joel I want to install Moshi hooks but they are closed. They have security implications...original β
16 Aug 2026
Qwen 3.8 27b just scored 4% higher than GLM 5.2 in a non-trivial private benchmark I am working...original β
16 Aug 2026
16 Aug 2026
15 Aug 2026
Lol now think how 27b would compare to the quantized opus 4.6 anthropic served us, instead of...original β
14 Aug 2026
14 Aug 2026