GRPO-4_40 / README.md

BRlkl

checkpoint 40-percent (step 38) (Trained with Unsloth)

c26adda verified about 1 month ago

preview code

raw

history blame contribute delete

558 Bytes

metadata

base_model: BRlkl/orchestrator-qwen3-4b-full
tags:
  - text-generation-inference
  - transformers
  - unsloth
  - qwen3
  - trl
  - grpo
license: apache-2.0
language:
  - en

Uploaded model

Developed by: BRlkl
License: apache-2.0
Finetuned from model : BRlkl/orchestrator-qwen3-4b-full

This qwen3 model was trained 2x faster with Unsloth