noneUsername/Cydonia-R1-24B-v4.1-W8A8
24B • Updated • 11
Username
noneUsername
AI & ML interests
None yet
Recent Activity
liked a model 3 days ago
perplexity-ai/pplx-decider-v1.3 reacted to sergiopaniego's post with 🚀 3 days ago
New post-training example: SFT on agent traces with TRL
→ A large model works as a coding agent in OpenCode, run through OpenEnv
→ The tasks it solves become the training data
→ A 2B model goes from 5.5% to 34.4% on held-out tasks
About 500 traces get most of the gain. Five times more add only 3 to 4 points, so SFT gets the agent working and RL has to do the rest
Full guide, with the models, traces and every rollout open:
https://github.com/huggingface/trl/tree/main/examples/sft_agent_traces liked a model 12 days ago
SeerRay-Lab/Xiaomi-Lens-0Organizations
None yet
models 79
noneUsername/Impish_Magic_24B-W8A8
24B • Updated • 9
noneUsername/Dolphin-Mistral-24B-Venice-Edition-W8A8
24B • Updated • 11
noneUsername/Qwen3-32B-abliterated-llm-compressor-AWQ
33B • Updated • 8
noneUsername/Mistral-Small-3.2-24B-Instruct-hf-W8A8
24B • Updated • 8 • 1
noneUsername/Homunculus-W8A8
12B • Updated • 8
noneUsername/Austral-24B-Winton-W8A8
24B • Updated • 5
noneUsername/Qwen3-32B-abliterated-awq
33B • Updated • 1.72k • 2
noneUsername/Magistral-Small-2506-W8A8
24B • Updated • 10
noneUsername/Broken-Tutu-24B-Unslop-v2.0-W8A8
24B • Updated • 11
datasets 0
None public yet