Update README.md

Upload plot.jpg

Update README.md

Update README.md

Update README.md

Update README.md

Update README.md

Update README.md

Delete mmlu_gpt2_step_1500.safetensors

Delete gpt2_step_60000(discard).safetensors

Rename gpt2_step_2000.safetensors to gpt2_step_52000(Smoltalk).safetensors

Checkpoint: Step 2000

Update README.md

Update README.md

Update README.md

Update README.md

Update README.md

Create README.md

Rename gpt2_step_50000_FIXED.safetensors to gpt2_step_50000.safetensors

✅ Fixed: Removed RMSNorm variance + used save_model for tied weights

Rename gpt2_step_50000.safetensors to gpt2_step_50000(bug).safetensors

Rename gpt2_step_60000.safetensors to gpt2_step_60000(discard).safetensors

Add MMLU fine-tuned checkpoint (step 1500, eval PPL 18.87)

Delete mmlu_gpt2_step_1000.safetensors

Add MMLU fine-tuned checkpoint (step 1000, eval PPL 13.62)

Upload gpt2_step_60000.safetensors with huggingface_hub

Upload gpt2_step_50000.safetensors with huggingface_hub

Upload gpt2_step_40000.safetensors with huggingface_hub

Upload gpt2_step_30000.safetensors with huggingface_hub

Upload gpt2_step_20000.safetensors with huggingface_hub

Upload gpt2_step_10000.safetensors with huggingface_hub

initial commit