36.7 GB

Ctrl+K

1 contributor

History: 124 commits

tekkmaven

Training status: step 352000

1ff8b12 verified 3 months ago

checkpoints
Checkpoint weights: step 352000 (1667MB) 3 months ago
configs
Upload configs/flint_40h.yaml 4 months ago
.gitattributes

1.52 kB
initial commit 4 months ago
README.md

9.66 kB
Update ML Intern artifact metadata 4 months ago
SCALING_GUIDE.md

3.43 kB
Upload SCALING_GUIDE.md 4 months ago
checkpointing.py

9.74 kB
Fix: push full model weights to Hub at hub_push_interval (not just status JSON)" 4 months ago
config.json

996 Bytes
Fix: config.json vocab_size 49216→49280 4 months ago
config.py

6.63 kB
Fix config.py default peak_lr to 1e-3 4 months ago
config.yaml

3.08 kB
Fix divergence: peak_lr 1e-3 → 3e-4, max_grad_norm 1.0 → 0.5 (batch=8 too small for high LR)" 4 months ago
convert_to_hf.py

5.82 kB
Upload convert_to_hf.py 4 months ago
data.py

16.3 kB
Fix data.py: handle edge cases, add fallback for failed datasets, fix orca-agentinstruct split names 4 months ago
data_pipeline.py

19.2 kB
Upload data_pipeline.py 4 months ago
evaluate.py

11.2 kB
Add evaluation script for benchmarking 4 months ago
export_hf.py

9.4 kB
Add model export utility (export_hf.py) for converting JAX weights to HF safetensors 4 months ago
muon.py

11.6 kB
Add Muon optimizer implementation (muon.py) 4 months ago
requirements.txt

525 Bytes
Upload requirements.txt 4 months ago
run_kaggle.py

3.89 kB
Always pass --resume so train.py checks Hub when local is empty" 4 months ago
train.py

18.2 kB
Auto-resume from Hub when local checkpoints are cleared by Kaggle\n\nNow --resume checks: local disk first → Hub download fallback → fresh start\nNo more lost progress across Kaggle sessions." 4 months ago
train_flint.py

34.4 kB
Upload train_flint.py 4 months ago
training_status.json

204 Bytes
Training status: step 352000 3 months ago