4 1 year ago

tools
ollama run owao/Deductive-Reasoning-Qwen-32B

Details

1 year ago

e86d20f4ce5a · 20GB ·

qwen2
·
32.8B
·
Q4_K_M
{ "num_ctx": 16384, "num_gpu": 65, "temperature": 0.6 }
{{- if .Messages }} {{- if or .System .Tools }}<|im_start|>system {{- if .System }} {{ .System }} {{

Readme

Credits to @DavidAU for his findings.

Used llama.cpp-b4837 for quantization.

Original model: OpenPipe/Deductive-Reasoning-Qwen-32B

Deductive-Reasoning-Qwen-32B

image/png

Deductive Reasoning Qwen 32B is a reinforcement fine-tune of Qwen 2.5 32B Instruct to solve challenging deduction problems from the Temporal Clue dataset, trained by OpenPipe!

Here are some additional resources to check out:

If you’re interested in training your own models with reinforcement learning or just chatting, feel free to reach out or email Kyle directly at kyle@openpipe.ai!