rapidfireai 0.0.1__py3-none-any.whl → 0.9.10__py3-none-any.whl

rapidfireai/ml/callbacks.py

@@ -0,0 +1,176 @@
+from typing import Callable, Dict, List, Optional
+
+import torch
+from datasets import Dataset
+from tqdm import tqdm
+from transformers import TrainerCallback, TrainerControl, TrainerState, TrainingArguments
+
+
+class GenerationMetricsCallback(TrainerCallback):
+    def __init__(
+        self,
+        tokenizer,
+        eval_dataset: Dataset,
+        generation_config: Optional[Dict] = None,
+        compute_metrics: Callable = None,
+        batch_size: int = 8,
+        mlflow_manager=None,
+        mlflow_run_id: str = None,
+        completed_steps: int = 0,
+    ):
+        self.tokenizer = tokenizer
+        self.eval_dataset = eval_dataset
+        self.compute_metrics = compute_metrics
+        self.batch_size = batch_size
+        self.generation_config = generation_config or {
+            "max_new_tokens": 128,
+            "temperature": 0.7,
+            "do_sample": True,
+            "top_p": 0.9,
+            "pad_token_id": tokenizer.pad_token_id,
+            "eos_token_id": tokenizer.eos_token_id,
+        }
+        self.mlflow_manager = mlflow_manager
+        self.mlflow_run_id = mlflow_run_id
+        self.completed_steps = completed_steps
+
+    def on_evaluate(self, args: TrainingArguments, state: TrainerState, control: TrainerControl, **kwargs):
+        model = kwargs.get("model")
+        if model is None:
+            return
+
+        metrics = self._compute_generation_metrics(model, state.global_step)
+
+        # Ensure metrics are added to log history
+        if hasattr(state, "log_history") and state.log_history:
+            state.log_history[-1].update(metrics)
+        else:
+            # If no log history exists, create a new entry
+            if not hasattr(state, "log_history"):
+                state.log_history = []
+            state.log_history.append(metrics)
+
+        for key, value in metrics.items():
+            if self.mlflow_manager:
+                self.mlflow_manager.log_metric(
+                    self.mlflow_run_id, key, value, step=self.completed_steps + state.global_step
+                )
+
+    def _prepare_data(self, eval_dataset: Dataset) -> tuple:
+        """Prepare batch data for generation"""
+        input_texts = []
+        references = []
+
+        for item in eval_dataset:
+            if isinstance(item, dict):
+                if "input" in item and "output" in item:
+                    input_text = item["input"]
+                    reference = item["output"]
+                elif "prompt" in item and "completion" in item:
+                    input_text = item["prompt"]
+                    reference = item["completion"][-1]["content"]
+                    input_text = self.tokenizer.apply_chat_template(input_text, tokenize=False)
+                else:
+                    continue
+
+                input_texts.append(input_text)
+                references.append(reference)
+
+        return input_texts, references
+
+    def _generate_batch(self, model, input_texts: List[str]) -> List[str]:
+        """Generate text for a batch of inputs"""
+        # Tokenize batch
+        inputs = self.tokenizer(
+            input_texts,
+            return_tensors="pt",
+            padding=True,
+            truncation=True,
+            max_length=512,  # Adjust based on your model's context length
+        ).to(model.device)
+
+        return inputs["input_ids"]
+
+    def _compute_generation_metrics(self, model, step: int) -> Dict[str, float]:
+        """Generate text and compute BLEU/ROUGE metrics with batch processing"""
+        model.eval()
+
+        # Determine evaluation samples
+        eval_size = len(self.eval_dataset)
+        indices = list(range(eval_size))
+
+        predictions = []
+        references = []
+
+        # Process in batches
+        input_texts, batch_references = self._prepare_data(self.eval_dataset)
+        input_ids = self._generate_batch(model, input_texts)
+        with torch.no_grad():
+            for i in tqdm(range(0, len(indices), self.batch_size), desc="Generating for metrics"):
+                input_ids_batch = input_ids[i : i + self.batch_size]
+                with torch.inference_mode(), torch.cuda.amp.autocast():
+                    outputs_batch = model.generate(input_ids_batch, **self.generation_config)
+                generated_texts = self.tokenizer.batch_decode(
+                    outputs_batch[:, input_ids_batch.shape[1] :], skip_special_tokens=True
+                )
+                predictions.extend(generated_texts)
+                references.extend(batch_references[i : i + self.batch_size])
+
+        # Compute metrics
+        metrics = {}
+        try:
+            if self.compute_metrics and predictions:
+                metrics = self.compute_metrics((predictions, references))
+        except Exception:
+            return {}
+
+        # Cleanup
+        del predictions, references
+        # if torch.cuda.is_available():
+        #     torch.cuda.empty_cache()
+
+        return metrics
+
+
+class MLflowLoggingCallback(TrainerCallback):
+    """Callback for logging metrics to MLflow during training"""
+
+    def __init__(
+        self,
+        mlflow_manager,
+        mlflow_run_id: str,
+        excluded_keys: list = None,
+        completed_steps: int = 0,
+        chunk_id: int = 0,
+        num_epochs_completed: int = 0,
+    ):
+        self.mlflow_manager = mlflow_manager
+        self.mlflow_run_id = mlflow_run_id
+        self.completed_steps = completed_steps
+        self.excluded_keys = excluded_keys or [
+            "step",
+            "epoch",
+        ]
+        self.chunk_id = chunk_id
+        self.num_epochs_completed = num_epochs_completed
+
+    def on_log(self, args: TrainingArguments, state: TrainerState, control: TrainerControl, logs=None, **kwargs):
+        """Called when the trainer logs metrics"""
+        if logs is not None:
+            for key, value in logs.items():
+                if isinstance(value, (int, float)) and key not in self.excluded_keys:
+                    try:
+                        self.mlflow_manager.log_metric(
+                            self.mlflow_run_id, key, value, step=self.completed_steps + state.global_step
+                        )
+                    except Exception as e:
+                        print(f"Warning: Failed to log metric {key} to MLflow: {e}")
+            self.mlflow_manager.log_metric(
+                self.mlflow_run_id, "chunk number", self.chunk_id, step=self.completed_steps + state.global_step
+            )
+            self.mlflow_manager.log_metric(
+                self.mlflow_run_id,
+                "num_epochs_completed",
+                self.num_epochs_completed,
+                step=self.completed_steps + state.global_step,
+            )