upgrade trl to 0.24.0 and liger to 0.6.3 (#3230)

* upgrade trl to 0.24.0

* fix reward collator init

* use newer DataCollatorForPreference instead

* DataCollatorForPreference doesn't use padding kwarg

* fix input id labels

* fix fbgemm-gpu version for pytorch versions

* tweak pinned deps

* transformers doesn't support hub 1.0 yet

* upgrade liger dep to 0.6.3

* set TORCH_CUDA_ARCH_LIST correctly

src/axolotl/core/builders/causal.py

@@ -12,7 +12,7 @@ from transformers import (
     EarlyStoppingCallback,
     Trainer,
 )
-from trl.trainer.utils import RewardDataCollatorWithPadding
+from trl.trainer.reward_trainer import DataCollatorForPreference
 
 from axolotl.core.builders.base import TrainerBuilderBase
 from axolotl.core.trainers import (
@@ -453,7 +453,7 @@ class HFCausalTrainerBuilder(TrainerBuilderBase):
                 BatchSamplerDataCollatorForSeq2Seq,
                 DataCollatorForSeq2Seq,
                 DataCollatorWithFlattening,
-                RewardDataCollatorWithPadding,
+                DataCollatorForPreference,
             ]
         ]
         collator_args = [self.tokenizer]
@@ -470,7 +470,10 @@ class HFCausalTrainerBuilder(TrainerBuilderBase):
             if kwargs and isinstance(kwargs, dict):
                 kwargs.update(collator_cls_and_kwargs[1])
         elif self.cfg.reward_model:
-            collator = RewardDataCollatorWithPadding
+            collator = DataCollatorForPreference
+            tokenizer = collator_args.pop(0)
+            kwargs["pad_token_id"] = tokenizer.pad_token_id
+            kwargs.pop("padding")
         elif use_batch_sampler_collator:
             # Use V2BatchSamplerDataCollatorForSeq2Seq for flex attention,
             # supported multipack models, or non-flash-attention llama

src/axolotl/prompt_strategies/bradley_terry/chat_template.py

@@ -71,10 +71,10 @@ class BTChatTemplateStrategy(ChatTemplateStrategy):
             ]
 
         return {
-            "input_ids_chosen": chosen_tokenized["input_ids"],
+            "chosen_input_ids": chosen_tokenized["input_ids"],
             "attention_mask_chosen": chosen_tokenized["attention_mask"],
             "labels_chosen": 1.0,
-            "input_ids_rejected": rejected_tokenized["input_ids"],
+            "rejected_input_ids": rejected_tokenized["input_ids"],
             "attention_mask_rejected": rejected_tokenized["attention_mask"],
             "labels_rejected": 0.0,
         }