Pinkstack
Pinkstack
AI & ML interests
LLMs, VLMS, fine-tuning, mid-training, pre-training, reasoning-LLMs, Small(ish) language models.
Recent Activity
updated
a model
about 6 hours ago
Pinkstackorg/Qwen3-Coder-pruned-20B-A3B
published
a model
about 6 hours ago
Pinkstackorg/Qwen3-Coder-pruned-20B-A3B
updated
a model
1 day ago
Pinkstack/Fijik1.5-2.6b-a380m-GGUF
Organizations
LLM(s) trained for Luau code generation, Luau is the coding language for Roblox. These models were continually pre-trained, then fine-tuned for chat.
Trained to reason properly.
Small but mighty model with incredible performance for it's small size.
First gen of the Pinkstack advanced reasoning models. Medium quality but better than the original models they are based on.
Higher quality reasoning model(s)
Advanced chat reasoning models trained on many different datasets, such as QwQ and Sonnet 3.5, these models excel in math, coding, explaining things.
-
Pinkstack/PARM-V2-QwQ-Qwen-2.5-o1-3B-GGUF
Text Generation • 3B • Updated • 258 • 7 -
Pinkstack/PARM-V1.5-base-QwQ-Qwen-2.5-o1-3B
Text Generation • 3B • Updated • 42 • 2 -
mradermacher/PARM-V1.5-QwQ-Qwen-2.5-o1-3B-VLLM-GGUF
3B • Updated • 725 • 2 -
Pinkstack/Parm-2-CoT-14B-16k-o1-QwQ
Text Generation • 15B • Updated • 58 • 7
Pinkstack general accuracy model(s). Created to hopefully be as accurate as possible.
Models which we made as an experiment.
-
Pinkstack/PARM-v2-ULTRA-o1-3B-vLLM
Text Generation • 3B • Updated • 14 • 1 -
Pinkstack/Superthoughts-lite-1.8B-experimental-o1-GGUF
Text Generation • 2B • Updated • 623 • 1 -
Pinkstack/Superthoughts-lite-1.8B-experimental-o1
Text Generation • 2B • Updated • 42 • 2 -
Pinkstack/Superthoughts-lite-v2-MOE-Llama3.2-experimental-0427
Text Generation • 4B • Updated • 17 • 1
LLM(s) trained for Luau code generation, Luau is the coding language for Roblox. These models were continually pre-trained, then fine-tuned for chat.
Higher quality reasoning model(s)
Trained to reason properly.
Advanced chat reasoning models trained on many different datasets, such as QwQ and Sonnet 3.5, these models excel in math, coding, explaining things.
-
Pinkstack/PARM-V2-QwQ-Qwen-2.5-o1-3B-GGUF
Text Generation • 3B • Updated • 258 • 7 -
Pinkstack/PARM-V1.5-base-QwQ-Qwen-2.5-o1-3B
Text Generation • 3B • Updated • 42 • 2 -
mradermacher/PARM-V1.5-QwQ-Qwen-2.5-o1-3B-VLLM-GGUF
3B • Updated • 725 • 2 -
Pinkstack/Parm-2-CoT-14B-16k-o1-QwQ
Text Generation • 15B • Updated • 58 • 7
Small but mighty model with incredible performance for it's small size.
Pinkstack general accuracy model(s). Created to hopefully be as accurate as possible.
First gen of the Pinkstack advanced reasoning models. Medium quality but better than the original models they are based on.
Models which we made as an experiment.
-
Pinkstack/PARM-v2-ULTRA-o1-3B-vLLM
Text Generation • 3B • Updated • 14 • 1 -
Pinkstack/Superthoughts-lite-1.8B-experimental-o1-GGUF
Text Generation • 2B • Updated • 623 • 1 -
Pinkstack/Superthoughts-lite-1.8B-experimental-o1
Text Generation • 2B • Updated • 42 • 2 -
Pinkstack/Superthoughts-lite-v2-MOE-Llama3.2-experimental-0427
Text Generation • 4B • Updated • 17 • 1