Beta
Explore
Marketplace
Neural Labs
Chat
Wallet
Docs
causal-prm-lasttoken-dream7b-gsm8k – AI Model by AnonyRepo | AlphaNeural AI
You can deploy this model and start earning money today!
AnonyRepo
/
causal-prm-lasttoken-dream7b-gsm8k
like
0
peft
process-reward-model
discrete-diffusion
gsm8k
lora
causal-attention
last-token-pool
Dream-org/Dream-v0-Instruct-7B
adapter
mit
us
Views
No views yet
Model card
Files and Versions
Community
API
Deploy
Causal PRM, last-token readout (Dream-7B, GSM8K)
LoRA adapter + reward head with causal attention + last-token pooling.
Base: Dream-org/Dream-v0-Instruct-7B (frozen)
Attention: causal
Readout: last non-MASK token pool
Training: 15,000 steps, seed 42
Snapshot accuracy at mask=0: 0.789 ± 0.011 (vs mean-pool causal 0.732)
See
bidir-prm-dream7b-gsm8k
for loading code.
