minimind/trainer
baihaochen 3233d9872c agentic 2026-06-12 15:11:31 +08:00
..
ds_config_pretrain.json fix: switch DeepSpeed config from bf16 to fp16, ZeRO-2 to ZeRO-1 (RTX 3090 compat) 2026-05-23 16:45:56 +08:00
ds_config_zero2.json feat: add SearchShortQA DeepSpeed pipeline 2026-06-02 00:59:46 +08:00
rollout_engine.py feat: add SearchShortQA DeepSpeed pipeline 2026-06-02 00:59:46 +08:00
search_shortqa_reward.py feat: add SearchShortQA DeepSpeed pipeline 2026-06-02 00:59:46 +08:00
train_agent.py feat: add SearchShortQA DeepSpeed pipeline 2026-06-02 00:59:46 +08:00
train_agent_deepspeed.py agentic 2026-06-12 15:11:31 +08:00
train_distillation.py feat: add SearchShortQA DeepSpeed pipeline 2026-06-02 00:59:46 +08:00
train_dpo.py feat: add SearchShortQA DeepSpeed pipeline 2026-06-02 00:59:46 +08:00
train_full_sft.py feat: add SearchShortQA DeepSpeed pipeline 2026-06-02 00:59:46 +08:00
train_full_sft_deepspeed.py agentic 2026-06-12 15:11:31 +08:00
train_grpo.py feat: add SearchShortQA DeepSpeed pipeline 2026-06-02 00:59:46 +08:00
train_lora.py feat: add SearchShortQA DeepSpeed pipeline 2026-06-02 00:59:46 +08:00
train_ppo.py feat: add SearchShortQA DeepSpeed pipeline 2026-06-02 00:59:46 +08:00
train_pretrain.py feat: add SearchShortQA DeepSpeed pipeline 2026-06-02 00:59:46 +08:00
train_pretrain_deepspeed.py fix: harden DeepSpeed SFT resume 2026-06-11 21:45:31 +08:00
train_search_grpo.py feat: add SearchShortQA DeepSpeed pipeline 2026-06-02 00:59:46 +08:00
train_search_grpo_deepspeed.py fix: keep only latest DeepSpeed checkpoint 2026-06-03 21:35:47 +08:00
train_tokenizer.py merge local changes 2026-05-12 11:18:45 +00:00
trainer_utils.py fix: harden DeepSpeed SFT resume 2026-06-11 21:45:31 +08:00