API: Add logprobs for chat completions

Adds chat completion logprob support using OAI's spec. Tokens are
not converted to tiktoken here since that will add an extra dependency
for no real reason.

Signed-off-by: kingbri <bdashore3@proton.me>

backends/exllamav2/model.py

@@ -520,6 +520,8 @@ class ExllamaV2Container:
                 joined_generation["token_probs"].update(
                     unwrap(generation.get("token_probs"), {})
                 )
+
+                # Include empty logprob dicts for index preservation
                 joined_generation["logprobs"].append(
                     unwrap(generation.get("logprobs"), {})
                 )