Fix: enforce positive memory_max_entries for in-memory cache

[mshr-h]

Passing invalid memory_max_entries to Cache can crash cachetool with unclear errors. This PR validates it upfront and fails with clearer messages.

Repros:

from dataclasses import dataclass
from dspy.clients.cache import Cache

@dataclass
class DummyResponse:
    message: str
    usage: dict

if __name__ == "__main__":
    memory_cache = Cache(
        enable_disk_cache=False,
        enable_memory_cache=True,
        disk_cache_dir="",
        disk_size_limit_bytes=0,
        memory_max_entries=1, # negative entry size
    )

    request = {"prompt": "Hello", "model": "openai/gpt-4o-mini", "temperature": 0.7}
    value = DummyResponse(message="This is a test response", usage={"prompt_tokens": 10, "completion_tokens": 20})
    memory_cache.put(request, value)  # -> ValueError: value too large

from dataclasses import dataclass
from dspy.clients.cache import Cache

@dataclass
class DummyResponse:
    message: str
    usage: dict

if __name__ == "__main__":
    memory_cache = Cache(
        enable_disk_cache=False,
        enable_memory_cache=True,
        disk_cache_dir="",
        disk_size_limit_bytes=0,
        memory_max_entries=None, # None
    )

    request = {"prompt": "Hello", "model": "openai/gpt-4o-mini", "temperature": 0.7}
    value = DummyResponse(message="This is a test response", usage={"prompt_tokens": 10, "completion_tokens": 20})
    memory_cache.put(request, value)  # -> TypeError: '>' not supported between instances of 'int' and 'NoneType'