add grpo

c7c477c7 · chenych · c7c477c7 · c7c477c7 · c7c477c7 · c7c477c7
Commit c7c477c7 authored Sep 24, 2025 by chenych
20 changed files
--- a/src/llamafactory/model/loader.py
+++ b/src/llamafactory/model/loader.py
--- a/src/llamafactory/model/model_utils/__init__.py
+++ b/src/llamafactory/model/model_utils/__init__.py
--- a/src/llamafactory/model/model_utils/attention.py
+++ b/src/llamafactory/model/model_utils/attention.py
--- a/src/llamafactory/model/model_utils/checkpointing.py
+++ b/src/llamafactory/model/model_utils/checkpointing.py
--- a/src/llamafactory/model/model_utils/embedding.py
+++ b/src/llamafactory/model/model_utils/embedding.py
--- a/src/llamafactory/model/model_utils/kv_cache.py
+++ b/src/llamafactory/model/model_utils/kv_cache.py
--- a/src/llamafactory/model/model_utils/liger_kernel.py
+++ b/src/llamafactory/model/model_utils/liger_kernel.py
--- a/src/llamafactory/model/model_utils/longlora.py
+++ b/src/llamafactory/model/model_utils/longlora.py
--- a/src/llamafactory/model/model_utils/misc.py
+++ b/src/llamafactory/model/model_utils/misc.py
--- a/src/llamafactory/model/model_utils/mod.py
+++ b/src/llamafactory/model/model_utils/mod.py
--- a/src/llamafactory/model/model_utils/moe.py
+++ b/src/llamafactory/model/model_utils/moe.py
--- a/src/llamafactory/model/model_utils/packing.py
+++ b/src/llamafactory/model/model_utils/packing.py
--- a/src/llamafactory/model/model_utils/quantization.py
+++ b/src/llamafactory/model/model_utils/quantization.py
--- a/src/llamafactory/model/model_utils/rope.py
+++ b/src/llamafactory/model/model_utils/rope.py
--- a/src/llamafactory/model/model_utils/unsloth.py
+++ b/src/llamafactory/model/model_utils/unsloth.py
--- a/src/llamafactory/model/model_utils/valuehead.py
+++ b/src/llamafactory/model/model_utils/valuehead.py
--- a/src/llamafactory/model/model_utils/visual.py
+++ b/src/llamafactory/model/model_utils/visual.py
--- a/src/llamafactory/model/patcher.py
+++ b/src/llamafactory/model/patcher.py
--- a/src/llamafactory/third_party/__init__.py
+++ b/src/llamafactory/third_party/__init__.py
--- a/src/llamafactory/third_party/muon/__init__.py
+++ b/src/llamafactory/third_party/muon/__init__.py