from typing import Any
import torch
from .base import BaseSampler
[docs]
class TemperatureSampler(BaseSampler):
"""
Temperature sampling adjusts the "sharpness" of the probability distribution.
- Low temperature (<1.0): More deterministic, picks high-probability tokens
- High temperature (>1.0): More random, flatter distribution
"""
[docs]
def __init__(self, temperature: float = 1.0):
"""
Initialize the temperature sampler.
Args:
temperature: Temperature value to control sampling sharpness
"""
super().__init__()
if temperature <= 0:
raise ValueError("Temperature must be positive")
self.temperature = temperature
def _apply_sampling(self, logits: torch.Tensor) -> torch.Tensor:
"""
Apply temperature scaling to the logits.
Args:
logits: Raw logits from the model
Returns:
torch.Tensor: Temperature-scaled logits
"""
return logits / self.temperature
[docs]
def sample(
self,
model: Any,
input_ids: torch.Tensor,
max_length: int = 100,
num_return_sequences: int = 1,
**kwargs,
) -> torch.Tensor:
"""
Generate text using temperature sampling.
Args:
model: The language model to sample from
input_ids: Input token IDs
max_length: Maximum length of the generated sequence
num_return_sequences: Number of sequences to return
**kwargs: Additional arguments
Returns:
torch.Tensor: Generated token IDs
"""
generated = input_ids.clone()
for _ in range(max_length):
logits = self._get_logits(model, generated)
logits = self._apply_sampling(logits)
next_tokens = self._sample_from_logits(logits, num_samples=1)
generated = torch.cat([generated, next_tokens], dim=1)
# Check if all sequences have generated an EOS token
if (next_tokens == model.config.eos_token_id).any():
break
return generated