Search results

Mem and compute inefficiency in fused_linear_cross_entropy_foward

Status: Open.

#1232 In linkedin/Liger-Kernel;

· tyler-romero opened on May 23, 2026

LigerORPOTrainer doesn't work when not using FSDP

Status: Open.

#1229 In linkedin/Liger-Kernel;

· Ola-Vish opened on May 19, 2026

Add fused Modulated RMSNorm for DiT-style scale/shift conditioning

Status: Open.

#1224 In linkedin/Liger-Kernel;

· yukiu00 opened on May 13, 2026

[RFC] Add Monitoring Interface for External Logging Backends

Status: Open.

#1213 In linkedin/Liger-Kernel;

· ParagEkbote opened on May 1, 2026

[New backend] cuTile backend for Liger kernels — performance data on B200 + integration interest

Status: Open.

#1205 In linkedin/Liger-Kernel;

· xjmxyt opened on Apr 27, 2026

Gemma4 text model support

Status: Open.

#1186 In linkedin/Liger-Kernel;

· ruilin-gif opened on Apr 6, 2026

Glm4vConfig object has no attribute 'hidden_size'

Status: Open.

#1152 In linkedin/Liger-Kernel;

· akowalsk opened on Mar 18, 2026

LigerFusedLinearGRPOLoss produces ~100x larger grad_norm than TRL's non-Liger-Kernel path due to missing vLLM IS correction and other differences

Status: Open.

#1082 In linkedin/Liger-Kernel;

· yukiu00 opened on Feb 7, 2026

Skip dW/dB computation in norm ops when weight/bias is frozen (LoRA/PEFT optimization)

Status: Open.

#1067 In linkedin/Liger-Kernel;

· yukiu00 opened on Feb 4, 2026

[RFC] Standardized Model Config for Benchmarks

Status: Open.

#1051 In linkedin/Liger-Kernel;

· Tcc0403 opened on Jan 29, 2026

Compatibility issue between liger-kernel and zero3: The size of tensor a (0) must match the size of tensor b (4096) at non-singleton dimension 1

Status: Open.

#1048 In linkedin/Liger-Kernel;

· Snowdar opened on Jan 29, 2026

[NPU] Improve the performance of the embedding for NPU

Status: Open.

#1036 In linkedin/Liger-Kernel;

· TianHao324 opened on Jan 21, 2026