Article
GGML and llama.cpp join HF to ensure the long-term progress of Local AI
- +4
ggerganov, ngxson, allozaur, lysandre, victor, julien-c
• • 510
Steffen Altmeier
LagOps
·
AI & ML interests
None yet
Recent Activity
new activity about 6 hours ago
UltimateIntent/HeatSeeker-284B-A13B-Lora:GGUF version? new activity 17 days ago
tarruda/DeepSeek-V4-Flash-GGUF:Quants for updated version? new activity 19 days ago
unsloth/DeepSeek-V4-Flash-0731-GGUF:Why do the K quants use I quants?Organizations
None yet
GGUF version?
➕ 1
2
#1 opened about 6 hours ago
by LagOps
Quants for updated version?
12
#2 opened 20 days ago
by LagOps
Why do the K quants use I quants?
#21 opened 19 days ago
by LagOps
PPQ(Q)/PPL(base) and KLD seems significantly worse than Step 3.5 Flash
👀 1
23
#2 opened 3 months ago
by 
tarruda
Big jump in loss?
7
#17 opened 4 months ago
by 
usernameSRSalreadyexists
Impressed by the model
#7 opened 4 months ago
by LagOps
upvoted an article 6 months ago
Looking forward to IQ4_XS!
🔥 3
24
#1 opened 6 months ago
by tarruda
Disabling/Reducing model reasoning
5
#22 opened 6 months ago
by 
Abdallah1997
commented on Transformers v5: Simple model definitions powering the AI ecosystem 9 months ago
Awesome work, great job of maintaining the library, it sure isn't easy with all the different architectures and advancements.
Any chance for an IQ3_K?
👍 2
34
#6 opened about 1 year ago
by 
original-el8
liked a
model 12 months ago
Qwen Image support
👍 4
3
#1 opened 12 months ago
by 
Sinaya
Awesome finetune
❤️ 4
4
#1 opened over 1 year ago
by 
sidran
`UD-Q4_K_XL` or `Q4_K_M`?
16
#6 opened over 1 year ago
by 
pootow
reacted to Undi95's post with 👍 over 1 year ago
Post
16370
Hi there!
If you want to create your own thinking model or do a better MistralThinker, I just uploaded my entire dataset made on Deepseek R1 and the axolotl config. (well I made them public)
Axolotl config : Undi95/MistralThinker-v1.1
The dataset : Undi95/R1-RP-ShareGPT3
You can also read all I did on those two discord screenshot from two days ago, I'm a little lazy to rewrite all kek.
Hope you will use them!
If you want to create your own thinking model or do a better MistralThinker, I just uploaded my entire dataset made on Deepseek R1 and the axolotl config. (well I made them public)
Axolotl config : Undi95/MistralThinker-v1.1
The dataset : Undi95/R1-RP-ShareGPT3
You can also read all I did on those two discord screenshot from two days ago, I'm a little lazy to rewrite all kek.
Hope you will use them!
- 6 replies