Improve timeout and retry handling with better error management and configuration options.

Acceptance Criteria:

• Add APITimeoutError exception handling
• Improve retry logic with exponential backoff
• Add configurable retry attempts
• Add timeout configuration per request
• Add comprehensive tests for timeout and retry scenarios
• Update documentation with timeout and retry examples

Technical Requirements:

• Implement proper timeout error handling
• Support configurable retry strategies
• Add exponential backoff for retries
• Maintain performance with retries

Example Usage:

llm = ChatGradient(
model="llama3.3-70b-instruct",
timeout=30.0,
max_retries=3
)

try:
response = llm.invoke("Hello world!")
except APITimeoutError:
print("Request timed out")

References:

Gradient SDK Timeout Documentation