XiaomiMiMo/MiMo-V2.6-Pro-RL
Text Generation • 1T • Updated • 28.3k • • 790
Yubo Wang
xiaoqianWX
AI & ML interests
AI-VIS team @Dapiya |
UW-Madison AOS
Recent Activity
liked a model about 1 month ago
XiaomiMiMo/MiMo-V2.6-Pro-RL published a dataset about 1 month ago
Dapiya/motis_public liked a model 6 months ago
deepseek-ai/DeepSeek-V4-ProOrganizations
liked a
model about 1 month ago
published a
dataset about 1 month ago
liked a
model 6 months ago
liked a
dataset 8 months ago
is HLE w/ CoT a typo?
#4 opened 8 months ago
by xiaoqianWX
liked a
model 9 months ago
liked a
model 12 months ago
liked a
model about 1 year ago
published a
dataset over 1 year ago
updated a
dataset over 1 year ago
updated a
model over 1 year ago
published a
model over 1 year ago
updated a
model over 1 year ago
published a
model over 1 year ago
updated a
model over 1 year ago
published a
model over 1 year ago
reacted to lewtun's post with 🔥 almost 2 years ago
Post
10,582
We are reproducing the full DeepSeek R1 data and training pipeline so everybody can use their recipe. Instead of doing it in secret we can do it together in the open!
🧪 Step 1: replicate the R1-Distill models by distilling a high-quality reasoning corpus from DeepSeek-R1.
🧠 Step 2: replicate the pure RL pipeline that DeepSeek used to create R1-Zero. This will involve curating new, large-scale datasets for math, reasoning, and code.
🔥 Step 3: show we can go from base model -> SFT -> RL via multi-stage training.
Follow along: https://github.com/huggingface/open-r1
🧪 Step 1: replicate the R1-Distill models by distilling a high-quality reasoning corpus from DeepSeek-R1.
🧠 Step 2: replicate the pure RL pipeline that DeepSeek used to create R1-Zero. This will involve curating new, large-scale datasets for math, reasoning, and code.
🔥 Step 3: show we can go from base model -> SFT -> RL via multi-stage training.
Follow along: https://github.com/huggingface/open-r1
- 5 replies
liked 2
models almost 2 years ago
vLLM on A100s
6
#41 opened almost 2 years ago
by 
fsaudm