List your company

Supervised Fine-Tuning on SFT (val)

1.3519Validation LossEmber
1.3054321.6190911.932752.246409Jul 1, 2026

Evaluation results

MethodLinks
EmberModel=pythia-2.8b, Params (M)=2800, Step=250, Optimizer State Precision=fp322026.071.35190.0008413,000
AdamWModel=pythia-2.8b, Params (M)=2800, Step=250, Optimizer State Precision=fp322026.071.3527—2,060
EmberModel=pythia-1.4b, Params (M)=1400, Step=250, Optimizer State Precision=fp322026.071.43060.0003409,000
AdamWModel=pythia-1.4b, Params (M)=1400, Step=250, Optimizer State Precision=fp322026.071.4309—1,650
EmberModel=gpt2-large, Params (M)=774, Step=250, Optimizer State Precision=fp322026.071.58880.0012201,000
AdamWModel=gpt2-large, Params (M)=774, Step=250, Optimizer State Precision=fp322026.071.59—515,000
EmberModel=gpt2-medium, Params (M)=350, Step=250, Optimizer State Precision=fp322026.071.7430.0037200,000
AdamWModel=gpt2-medium, Params (M)=350, Step=250, Optimizer State Precision=fp322026.071.7467—412,000
EmberModel=gpt2, Params (M)=124, Step=250, Optimizer State Precision=fp322026.072.03810.0075199,000
AdamWModel=gpt2, Params (M)=124, Step=250, Optimizer State Precision=fp322026.072.0456—309,000
EmberModel=distilgpt2, Params (M)=82, Step=250, Optimizer State Precision=fp322026.072.35780.0062199,000
AdamWModel=distilgpt2, Params (M)=82, Step=250, Optimizer State Precision=fp322026.072.364—309,000
AdamWModel=pythia-160m, Params (M)=160, Step=250, Optimizer State Precision=fp322026.072.4782—618,000
EmberModel=pythia-160m, Params (M)=160, Step=250, Optimizer State Precision=fp322026.072.51360.0354399,000