Reinforcement Learning
Transformers
English
post-training
distillation
agentic-coding
composer-2.5
cursor
kimi-k2
grpo
dapo
diloco
openenv
trl
verl
research
methodology
Instructions to use Codeseys/composer-replication-framework with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Codeseys/composer-replication-framework with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("Codeseys/composer-replication-framework", dtype="auto") - Notebooks
- Google Colab
- Kaggle
| { | |
| "status": "PASS", | |
| "dtype": "torch.bfloat16", | |
| "n_steps": 30, | |
| "sdpo_fired_nonzero": true, | |
| "max_sdpo_kl": 0.185546875, | |
| "loss_trend_down": true, | |
| "all_finite": true, | |
| "loss_first": 4.726175785064697, | |
| "loss_last": 0.0050478726625442505, | |
| "loss_curve": [ | |
| 4.7262, | |
| 2.9752, | |
| 1.8887, | |
| 1.1304, | |
| 0.6017, | |
| 0.3641, | |
| 0.1361, | |
| 0.037, | |
| 0.0142, | |
| 0.0077, | |
| 0.0061, | |
| 0.0055, | |
| 0.0053, | |
| 0.0053, | |
| 0.0052, | |
| 0.0052, | |
| 0.0052, | |
| 0.0051, | |
| 0.0051, | |
| 0.0051, | |
| 0.0051, | |
| 0.0051, | |
| 0.0051, | |
| 0.0051, | |
| 0.0051, | |
| 0.0051, | |
| 0.0051, | |
| 0.005, | |
| 0.005, | |
| 0.005 | |
| ], | |
| "sdpo_curve": [ | |
| 0.0801, | |
| 0.0757, | |
| 0.0771, | |
| 0.0903, | |
| 0.1387, | |
| 0.1855, | |
| 0.1445, | |
| 0.1113, | |
| 0.0679, | |
| 0.0574, | |
| 0.0381, | |
| 0.0151, | |
| 0.009, | |
| 0.0066, | |
| 0.0065, | |
| 0.0052, | |
| 0.0048, | |
| 0.0038, | |
| 0.0031, | |
| 0.0021, | |
| 0.0026, | |
| 0.0029, | |
| 0.0016, | |
| 0.0013, | |
| 0.0021, | |
| 0.0018, | |
| 0.0014, | |
| 0.0011, | |
| 0.0007, | |
| 0.0015 | |
| ], | |
| "run_date": "2026-05-29", | |
| "gpu": "A10G", | |
| "model": "Qwen/Qwen2.5-0.5B-Instruct" | |
| } |