[feat] Support DAPO (#6263)

* update help information

* update style

* fix

* minor fix

* support PP training

* add pp support

* remove unused code

* address conversation

* fix memory leakage support tp+pp

* move empty cache

* move empty cache

* add DAPO support

* remove format reward

* fix filtering, still buggy

* small fix

* add DAPO support

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* tested multi-node training; fix bind_batch bug

* fix conversation; support sleep mode

* support reusing excessive samples

* add dynamic batching control flag

* add dynamic batching control flag

* refactored

* fix logging

---------

Co-authored-by: Tong Li <[email protected]>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

Author: 3 people

applications/ColossalChat/coati/distributed/consumer.py

@@ -16,7 +16,7 @@
 from colossalai.utils import get_current_device
 
 from .comm import ray_broadcast_tensor_dict
-from .utils import bind_batch, post_recv, unbind_batch
+from .utils import bind_batch, pad_batch, post_recv, unbind_batch
 
 
 class BaseConsumer:
@@ -33,7 +33,7 @@ def __init__(
         batch_size: int,
         model_config: Dict[str, Any],
         plugin_config: Dict[str, Any],
-        microbatch_size: int = 1,
+        minibatch_size: int = 1,
         save_interval: int = 100,
         save_dir: str = "./model",
     ):
@@ -46,11 +46,11 @@ def __init__(
         self.num_update_per_episode = num_update_per_episode
         self.num_recv_per_update = num_recv_per_update
         self.batch_size = batch_size
-        self.microbatch_size = microbatch_size
+        self.minibatch_size = minibatch_size
         self.save_interval = save_interval
         self.save_dir = save_dir
-        assert batch_size % microbatch_size == 0, "batch_size should be divisible by microbatch_size"
-        self.num_microbatches = batch_size // microbatch_size
+        assert batch_size % minibatch_size == 0, "batch_size should be divisible by microbatch_size"
+        self.num_microbatches = batch_size // minibatch_size
 
         self.model_config = model_config
         self.plugin_config = plugin_config
@@ -67,7 +67,7 @@ def setup(self) -> None:
 
         plugin_config = dict(tp_size=1, pp_size=1, precision="bf16", zero_stage=2)
         if self.plugin_config.get("pp_size", 1) > 1 and "num_microbatches" not in self.plugin_config:
-            plugin_config["microbatch_size"] = self.microbatch_size
+            plugin_config["microbatch_size"] = self.minibatch_size
         plugin_config.update(self.plugin_config)
         self.plugin = HybridParallelPlugin(**plugin_config)
         self.booster = Booster(plugin=self.plugin)
@@ -105,18 +105,26 @@ def loop(self) -> None:
                                     )
                                 )
                             )
-                        while len(self.buffer) >= self.dp_size * self.microbatch_size:
+                        while len(self.buffer) >= self.dp_size * self.minibatch_size:
                             batches = self.buffer[
-                                self.dp_rank * self.microbatch_size : (self.dp_rank + 1) * self.microbatch_size
+                                self.dp_rank * self.minibatch_size : (self.dp_rank + 1) * self.minibatch_size
                             ]
-                            self.buffer = self.buffer[self.dp_size * self.microbatch_size :]
+                            batch = pad_batch(
+                                batches
+                            )  # when `imbs` is smaller than `tMbs`, samples may have differ in size, need to pad before stacking
                             batch = bind_batch(batches)
                             batch = post_recv(batch)
-                            loss = self.step(i, **batch)
+                            loss, num_excessive_prompts = self.step(i, pbar, **batch)
+                            self.buffer = (
+                                self.buffer[
+                                    (self.dp_rank + 1) * self.minibatch_size
+                                    - num_excessive_prompts : (self.dp_rank + 1) * self.minibatch_size
+                                ]
+                                + self.buffer[self.dp_size * self.minibatch_size :]
+                            )
                             if loss is not None:
                                 pbar.set_postfix({"loss": loss})
                             i += 1
-                    assert len(self.buffer) == 0
                     if self.lr_scheduler is not None:
                         self.lr_scheduler.step()
                     if (step + 1) % self.save_interval == 0 or (step + 1) == self.num_update_per_episode:
@@ -154,7 +162,9 @@ def __init__(
         batch_size,
         model_config,
         plugin_config,
-        microbatch_size=1,
+        minibatch_size=1,
+        save_interval: int = 100,
+        save_dir="./model",
     ):
         super().__init__(
             num_producers,
@@ -168,7 +178,7 @@ def __init__(
             batch_size,
             model_config,
             plugin_config,
-            microbatch_size,
+            minibatch_size,
         )
         path = model_config.pop("path")
         self.model = AutoModelForCausalLM.from_pretrained(path, **model_config)
@@ -181,7 +191,7 @@ def setup(self):
         super().setup()
         self.model, self.optimizer, *_ = self.booster.boost(self.model, self.optimizer)
 
-    def step(self, step_idx: int, **kwargs) -> Optional[float]:
+    def step(self, step_idx: int, pbar: Any, **kwargs) -> Optional[float]:
         labels = kwargs["input_ids"].clone()
         labels[kwargs["attention_mask"] == 0] = -100
         kwargs["labels"] = labels