Fixed the Q-learning update rule in qlearning.md(Issue no #2649)

[DEVANSH-GAJJAR]

Description of changes:

• Updated the Q-learning update equation to use td_target and td_error for clarity.
• Matches the standard Q-learning formula.
• Fixes issue Error in chapter 17, section 3 "Q-learning", equation 17.3.3 and 17.3.4. #2649.

[DEVANSH-GAJJAR]

Hello There , I’ve fixed the Q-learning update rule as per issue #2649. It’s a small change inside qlearning.md, so the CI failures on JAX, PyTorch, TensorFlow, and MXNet builds may be unrelated. Let me know if there’s anything I should fix or adjust.