Beinsezii/Krea-2-Turbo-Projector-Scale-LoRA-Diffusers
Updated • 842 • 22
DepictDev
Depict
·
AI & ML interests
None yet
Recent Activity
liked a model 1 day ago
Beinsezii/Krea-2-Turbo-Projector-Scale-LoRA-Diffusers liked a model 2 days ago
AlperKTS/Krea2_FP8 liked a model 6 days ago
osunlp/QUEST-35B-RLOrganizations
None yet
liked a
model 1 day ago
liked a
model 2 days ago
liked a
model 6 days ago
upvoted a paper 9 days ago
liked a
model 25 days ago
liked 4
models about 1 month ago
reacted to salma-remyx's post with 🚀 about 1 month ago
Post
5130
Just trained a 2B coding model to rank candidate AI/ML research ideas against the implicit preferences in a code repository's merge history.
The training data comes from a Gaussian Process fit on the accumulated dispositions in VQASynth, where each PR against a deployed project yields a pairwise comparison between the feature branch preferred and the baseline at main.
The GP scores candidate papers to synthesize preference pairs, and DPO with LoRA bakes the ranking pipeline into the model's weights.
After 1 epoch the model reaches 87.4% reward accuracy on the held-out eval split against 92.3% on training, consistent with learning the task without overfitting.
Now, I'm scaling the pipeline to thousands of repos for a generalization test.
Dataset: remyxai/mhpd-dpo-v0
Model: remyxai/mhpd-dpo-qwen3.5-2b-vqasynth
Substack: https://remyxai.substack.com/p/the-ai-pm
The training data comes from a Gaussian Process fit on the accumulated dispositions in VQASynth, where each PR against a deployed project yields a pairwise comparison between the feature branch preferred and the baseline at main.
The GP scores candidate papers to synthesize preference pairs, and DPO with LoRA bakes the ranking pipeline into the model's weights.
After 1 epoch the model reaches 87.4% reward accuracy on the held-out eval split against 92.3% on training, consistent with learning the task without overfitting.
Now, I'm scaling the pipeline to thousands of repos for a generalization test.
Dataset: remyxai/mhpd-dpo-v0
Model: remyxai/mhpd-dpo-qwen3.5-2b-vqasynth
Substack: https://remyxai.substack.com/p/the-ai-pm
liked 6
models about 1 month ago
llmfan46/Qwen3.6-27B-uncensored-heretic-v2-Native-MTP-Preserved-GGUF
Image-Text-to-Text • Updated • 106k • 139
llmfan46/Gemma-4-Gembrain-31B-it-uncensored-heretic-GGUF
31B • Updated • 4.61k • 12
bytedance-research/Lance
Any-to-Any • Updated • 1.7k • 1.05k
numind/NuExtract-2.0-8B
Image-Text-to-Text • 8B • Updated • 23.5k • 63
Jackrong/Qwopus3.5-9B-Coder-GGUF
Image-Text-to-Text • 9B • Updated • 75.5k • 288
Nimbz/Gemma-4-Gembrain-31B
31B • Updated • 90 • 41
liked 4
models about 2 months ago
SulphurAI/Sulphur-2-base
Text-to-Video • 9B • Updated • 908k • 1.77k
llmfan46/Gemma-4-Garnet-31B-it-uncensored-heretic-GGUF
Text Generation • 31B • Updated • 1.36k • 9
Qwen/WebWorld-32B
Text Generation • 677k • Updated • 664 • 68
ibm-granite/granite-switch-4.1-8b-preview
Text Generation • 10B • Updated • 1.06k • 27