Downloads · 30 days
4
12% of all-time downloads
Mahesh111000/env2-k-qwen3-32b-step125
env2-k-qwen3-32b-step125 is a reinforcement learning model from Mahesh111000. Use it for the reinforcement learning task on the model card, and read the license before you ship it in a product. The card lists the license as apache-2.0.
Qwen3-32B after 125 steps of RL on the env-2 K-expression task (the checkpoint evaluated in the paper). Full merged weights: LoRA rank 32 / alpha 64 merged into the base in fp32 and saved in fp32 so the RL delta is pr…
Downloads · 30 days
4
12% of all-time downloads
All-time downloads
33
Public
Parameters
32.8B
131 GB on disk
Likes
0
Public
Click a slice to open those files.
.safetensors131 GB · 100%
From the Hugging Face model README
Qwen3-32B after 125 steps of RL on the env-2 K-expression task (the checkpoint evaluated in the paper). Full merged weights: LoRA rank 32 / alpha 64 merged into the base in fp32 and saved in fp32 so the RL delta is preserved exactly (bf16 rounding of merged weights drops part of it). Thinking mode.