-
contest-recipe (bigram init, no ramps, no lr warmup, 1,000 steps) model states at 0/15/250/500/1000
-
Upload README.md with huggingface_hub
-
Upload checkpoints/40GB-A100_d12_walkthrough/state_step000250.pt with huggingface_hub
-
Upload checkpoints/40GB-A100_d12_walkthrough/state_step000015.pt with huggingface_hub
-
Upload README.md with huggingface_hub
-
Upload checkpoints/40GB-A100_d12_walkthrough/state_step000250.pt with huggingface_hub
-
Upload checkpoints/40GB-A100_d12_walkthrough/state_step000015.pt with huggingface_hub
-
initial commit