support dpo-ftx

Former-commit-id: 86dfa04f9821556019fa777106787f73eb70b452

src/llmtuner/train/dpo/trainer.py

@@ -16,10 +16,11 @@ class CustomDPOTrainer(DPOTrainer):
     def __init__(
         self,
         beta: float,
+        loss_type: Literal["sigmoid", "hinge"],
+        ftx_gamma: float,
         model: Union["PreTrainedModel", torch.nn.Module],
         ref_model: Optional[Union["PreTrainedModel", torch.nn.Module]] = None,
         disable_dropout: Optional[bool] = True,
-        loss_type: Optional[Literal["sigmoid", "hinge"]] = "sigmoid",
         **kwargs
     ):
         if disable_dropout:
@@ -34,6 +35,8 @@ class CustomDPOTrainer(DPOTrainer):
         self.label_pad_token_id = IGNORE_INDEX
         self.padding_value = 0
         self.beta = beta
+        self.label_smoothing = 0
+        self.ftx_gamma = ftx_gamma
         self.loss_type = loss_type
         self._stored_metrics = defaultdict(lambda: defaultdict(list))
 
@@ -51,10 +54,28 @@ class CustomDPOTrainer(DPOTrainer):
             else:
                 self.ref_model = self.accelerator.prepare_model(self.ref_model, evaluation_mode=True)
 
+    def sft_loss(
+        self,
+        chosen_logits: torch.FloatTensor,
+        chosen_labels: torch.LongTensor
+    ) -> torch.Tensor:
+        r"""
+        Computes supervised cross-entropy loss of given labels under the given logits.
+
+        Returns:
+            A tensor of shape (batch_size,) containing the cross-entropy loss of each samples.
+        """
+        all_logps = self._get_batch_logps(
+            chosen_logits,
+            chosen_labels,
+            average_log_prob=True
+        )
+        return -all_logps
+
     def concatenated_forward(
         self,
-        model: Optional[torch.nn.Module] = None,
-        batch: Optional[Dict[str, torch.Tensor]] = None
+        model: "PreTrainedModel",
+        batch: Dict[str, torch.Tensor]
     ) -> Tuple[torch.FloatTensor, torch.FloatTensor, torch.FloatTensor, torch.FloatTensor]:
         batch_copied = BatchEncoding({k: v.detach().clone() for k, v in batch.items()}) # avoid error
 
@@ -73,3 +94,61 @@ class CustomDPOTrainer(DPOTrainer):
         chosen_logps, rejected_logps = all_logps.split(batch_size, dim=0)
         chosen_logits, rejected_logits = all_logits.split(batch_size, dim=0)
         return chosen_logps, rejected_logps, chosen_logits, rejected_logits
+
+    def get_batch_metrics(
+        self,
+        model: "PreTrainedModel",
+        batch: Dict[str, torch.Tensor],
+        train_eval: Optional[Literal["train", "eval"]] = "train"
+    ) -> Tuple[torch.Tensor, Dict[str, torch.Tensor]]:
+        r"""
+        Computes the DPO loss and other metrics for the given batch of inputs for train or test.
+        """
+        metrics = {}
+        (
+            policy_chosen_logps,
+            policy_rejected_logps,
+            policy_chosen_logits,
+            policy_rejected_logits,
+        ) = self.concatenated_forward(model, batch)
+        with torch.no_grad():
+            if self.ref_model is None:
+                with self.accelerator.unwrap_model(self.model).disable_adapter():
+                    (
+                        reference_chosen_logps,
+                        reference_rejected_logps,
+                        _,
+                        _,
+                    ) = self.concatenated_forward(self.model, batch)
+            else:
+                (
+                    reference_chosen_logps,
+                    reference_rejected_logps,
+                    _,
+                    _,
+                ) = self.concatenated_forward(self.ref_model, batch)
+
+        losses, chosen_rewards, rejected_rewards = self.dpo_loss(
+            policy_chosen_logps,
+            policy_rejected_logps,
+            reference_chosen_logps,
+            reference_rejected_logps,
+        )
+        if self.ftx_gamma > 1e-6:
+            batch_size = batch["input_ids"].size(0) // 2
+            chosen_labels, _ = batch["labels"].split(batch_size, dim=0)
+            losses += self.ftx_gamma * self.sft_loss(policy_chosen_logits, chosen_labels)
+
+        reward_accuracies = (chosen_rewards > rejected_rewards).float()
+
+        prefix = "eval_" if train_eval == "eval" else ""
+        metrics[f"{prefix}rewards/chosen"] = chosen_rewards.cpu().mean()
+        metrics[f"{prefix}rewards/rejected"] = rejected_rewards.cpu().mean()
+        metrics[f"{prefix}rewards/accuracies"] = reward_accuracies.cpu().mean()
+        metrics[f"{prefix}rewards/margins"] = (chosen_rewards - rejected_rewards).cpu().mean()
+        metrics[f"{prefix}logps/rejected"] = policy_rejected_logps.detach().cpu().mean()
+        metrics[f"{prefix}logps/chosen"] = policy_chosen_logps.detach().cpu().mean()
+        metrics[f"{prefix}logits/rejected"] = policy_rejected_logits.detach().cpu().mean()
+        metrics[f"{prefix}logits/chosen"] = policy_chosen_logits.detach().cpu().mean()
+
+        return losses.mean(), metrics