sphiratrioth666/SillyTavern-Presets-Sphiratrioth
Updated
•
68
Yamam
Herman555
AI & ML interests
None yet
Recent Activity
liked
a model
13 days ago
sphiratrioth666/SillyTavern-Presets-Sphiratrioth
Organizations
None yet
Herman555's activity
liked
a
model
13 days ago
New activity in
DavidAU/Maximizing-Model-Performance-All-Quants-Types-And-Full-Precision-by-Samplers_Parameters
28 days ago
Model classes???
3
#2 opened 28 days ago
by
Herman555
Koboldcpp customization request
7
#1 opened 30 days ago
by
FrescoHF
Brief feedback.
6
#2 opened 4 months ago
by
Herman555
my five cents on Stheno 3.4
7
#1 opened 4 months ago
by
CynicalSpore
reacted to
Undi95's
post with ❤️
5 months ago
Post
11877
Hello there,
New model released, my goal was to try finetune on the last Llama-3.1-8B-Instruct but not a small train, I wanted to do something useful.
One of the rare model that I didn't made for RP, or in the goal to uncensor it (but I did anyway kek).
The model was trained on 9M Claude conversations ONLY, giving him another writting style.
Undi95/Meta-Llama-3.1-8B-Claude > OG release fp32, it's the epoch 2
Undi95/Meta-Llama-3.1-8B-Claude-bf16 > Base model resharded in bf16 waiting for available quant without issues
Since it's frustrating to be censored using a local model, orthogonal activation steering was used, trying to force the model to never refuse a prompt.
Undi95/Meta-Llama-3.1-8B-Claude-68fail-3000total > Uncensored model, refuse 68 times on 3000 toxic prompt
Undi95/Meta-Llama-3.1-8B-Claude-39fail-3000total > Uncensored model, refuse 39 times on 3000 toxic prompt
It still refuse some prompt but the majority of them is uncensored. OAS can make a model more dumb or make the base perplexity go higher, so I didn't snipe for 0 refusal.
I don't do non-RP model a lot so any feedback is welcome, I would like to re-use this base for some others future project if needed.
New model released, my goal was to try finetune on the last Llama-3.1-8B-Instruct but not a small train, I wanted to do something useful.
One of the rare model that I didn't made for RP, or in the goal to uncensor it (but I did anyway kek).
The model was trained on 9M Claude conversations ONLY, giving him another writting style.
Undi95/Meta-Llama-3.1-8B-Claude > OG release fp32, it's the epoch 2
Undi95/Meta-Llama-3.1-8B-Claude-bf16 > Base model resharded in bf16 waiting for available quant without issues
Since it's frustrating to be censored using a local model, orthogonal activation steering was used, trying to force the model to never refuse a prompt.
Undi95/Meta-Llama-3.1-8B-Claude-68fail-3000total > Uncensored model, refuse 68 times on 3000 toxic prompt
Undi95/Meta-Llama-3.1-8B-Claude-39fail-3000total > Uncensored model, refuse 39 times on 3000 toxic prompt
It still refuse some prompt but the majority of them is uncensored. OAS can make a model more dumb or make the base perplexity go higher, so I didn't snipe for 0 refusal.
I don't do non-RP model a lot so any feedback is welcome, I would like to re-use this base for some others future project if needed.
- 4 replies
liked
a
model
7 months ago
Feedback
14
#1 opened 7 months ago
by
TravelingMan
replied to
Lewdiculous's
post
8 months ago
Yes, that's the one. Thank you so much!. You're genuinely awesome in every way.
replied to
Lewdiculous's
post
8 months ago
Can you also do the same for TheSpice model?, I still find it to be the best for roleplay currently.
reacted to
Lewdiculous's
post with ❤️👍
8 months ago
Post
43907
Updated: Lumimaid and TheSpice-v0.8.3
I have uploaded version 2 (v2) files for the Llama-3-Lumimaid-8B-v0.1-OAS GGUF Imatrix quants.
[model] Lewdiculous/Llama-3-Lumimaid-8B-v0.1-OAS-GGUF-IQ-Imatrix
You can recognize the new files by their
Imatrix data was generated from the FP16 and conversions directly from the BF16.
Hopefully avoiding any losses in the model conversion, as has been the recently discussed topic on Llama-3 and GGUF lately.
This is more disk and compute intensive so lets hope we get GPU inference support for BF16 models in llama.cpp.
If you are able to test them and noticed any issues compared to the original quants, let me know in the corresponding discussions.
---
Additionally, L3-TheSpice-8b-v0.8.3 GGUF Imatrix quants were also updated.
[model] Lewdiculous/L3-TheSpice-8b-v0.8.3-GGUF-IQ-Imatrix
I have uploaded version 2 (v2) files for the Llama-3-Lumimaid-8B-v0.1-OAS GGUF Imatrix quants.
[model] Lewdiculous/Llama-3-Lumimaid-8B-v0.1-OAS-GGUF-IQ-Imatrix
You can recognize the new files by their
v2 prefix.Imatrix data was generated from the FP16 and conversions directly from the BF16.
Hopefully avoiding any losses in the model conversion, as has been the recently discussed topic on Llama-3 and GGUF lately.
This is more disk and compute intensive so lets hope we get GPU inference support for BF16 models in llama.cpp.
If you are able to test them and noticed any issues compared to the original quants, let me know in the corresponding discussions.
---
Additionally, L3-TheSpice-8b-v0.8.3 GGUF Imatrix quants were also updated.
[model] Lewdiculous/L3-TheSpice-8b-v0.8.3-GGUF-IQ-Imatrix
-
- 7 replies
The prompt results in censored outputs.
13
#4 opened 8 months ago
by
Herman555
Amazing at roleplay
5
#1 opened 8 months ago
by
AliceThirty
Seems to be broken.
5
#1 opened 8 months ago
by
Herman555
Llama 3 is out!
2
#6 opened 8 months ago
by
Herman555
Prompt format.
3
#2 opened 9 months ago
by
Herman555
Are we 75% there?
4
#5 opened 10 months ago
by
Lewdiculous
Great model.
11
#2 opened 10 months ago
by
Herman555