Spaces:

zhan1206
/

fusion-llm-demo

Running

App Files Files Community

fusion-llm-demo / train

Ctrl+K

Ctrl+K

3 contributors

History: 30 commits

zhan1206

fix: F4-1/2/3 — three crash-on-run bugs

bb40eba about 1 month ago

dpo_finetune.py

14.6 kB
v15: fix remaining audit issues (F1/F2/S1/M3/M8/doc) about 1 month ago
e2e_validation.py

9.99 kB
v2.2.0: end-to-end pipeline validation + GRPO no-tokenizer fallback about 1 month ago
full_finetune.py

12 kB
v17: fix Thinking Dial training pipeline (N10/N11/N12/N13/N14/N15/N16/N17) about 1 month ago
gsm8k_eval.py

12.2 kB
feat: GSM8K reward function + GRPO evaluation pipeline (v2.1.1) about 1 month ago
lora_finetune.py

16.4 kB
fix: audit v2 P0/P1 fixes (D1/D2/D3/D7/D9/D11/D14) about 1 month ago
model_utils.py

2.73 kB
v17: fix Thinking Dial training pipeline (N10/N11/N12/N13/N14/N15/N16/N17) about 1 month ago
test_train_mini.py

3.38 kB
Train: Add minimal training test (1-2 steps, loss decreases) about 1 month ago
train_10steps.py

3.55 kB
Train: Add 10-step training test (loss decreases steadily) about 1 month ago
train_grpo.py

7.54 kB
fix: F4-1/2/3 — three crash-on-run bugs about 1 month ago
train_mini.py

12.9 kB
fix: comprehensive audit fixes - CRITICAL/HIGH/MEDIUM/LOW about 1 month ago
train_real.py

6.46 kB
fix: comprehensive audit fixes - CRITICAL/HIGH/MEDIUM/LOW about 1 month ago
training_optimizations.py

6.91 kB
Feat: Add AMP trainer and gradient accumulation for training optimization about 1 month ago