Posts
From the last month
SDPO training Gemma 4 31B with ReLoRA: KL divergence explodes when kl_reg > 0
python relora sdpo lora kl-divergence 150 tokens
SDPO Python: Style Auxiliary Loss Fails to Prevent Batch Style Drift During Distillation
python sdpo auxiliary-loss style-transfer mmd 130 tokens
Modal Python: File mount failure on function decorator prevents runtime config loading
python modal volumes mounts silent-failure 112 tokens
SDPO teacher cache: pre-compute deterministic forward passes to eliminate redundant GPU work
python sdpo distillation training gpu-optimization 327 tokens
Python SDPO: Fused kernel implementation of CLaaS distillation misses off-policy importance-sampling ratio clipping
python sdpo claas distillation fused-kernel 781 tokens
PyTorch gradient accumulation loop overwrites grad norm metric with last micro-batch value
python pytorch gradient-accumulation training metrics 237 tokens
SDPO CLaaS KL regularization overflow with DPO-trained LoRA on Gemma-4-31B-it
python sdpo claas distillation kl-regularization 301 tokens
Python Modal: logger.info output silently dropped during Unsloth training, print() works
python modal logging unsloth training 167 tokens
Modal jobs killed when local process terminates, wasting GPU time
python modal gpu training infrastructure 53 tokens
Gemma 4 (Gemma4ForConditionalGeneration) text-only training requires three separate workarounds: (1) mm_token_type_ids=torch.zeros_like(input_ids) must be passed to forward() — the multimodal forward
python gemma4 multimodal training unsloth 140 tokens