fix(streaming): word-sliced cache replay for streaming completions
#25670Comparing
michelligabriele:fix/cached-response-streaming-cadence
(199c7a3
) withmain
(72a461b
)0%
- Untouched
- 16
Benchmarks
16 total
test_get_model_info_with_provider
tests/benchmarks/test_benchmarks.py+1%
81.1 µs80 µs
test_token_counter_simple_message
tests/benchmarks/test_benchmarks.py+1%
238.8 µs236.1 µs
test_get_model_info_anthropic
tests/benchmarks/test_benchmarks.py+1%
79.8 µs79 µs
test_token_counter_raw_text
tests/benchmarks/test_benchmarks.py+1%
184.7 µs183.1 µs
test_token_counter_with_tools
tests/benchmarks/test_benchmarks.py0%
415.5 µs413.6 µs
test_cost_per_token_openai
tests/benchmarks/test_benchmarks.py0%
542 µs539.6 µs
test_get_llm_provider_azure
tests/benchmarks/test_benchmarks.py0%
144.1 µs143.6 µs
test_token_counter_long_content
tests/benchmarks/test_benchmarks.py0%
1.7 ms1.7 ms
test_get_llm_provider_with_prefix
tests/benchmarks/test_benchmarks.py0%
137 µs136.7 µs
test_get_model_info_openai
tests/benchmarks/test_benchmarks.py0%
81 µs80.9 µs
test_cost_per_token_anthropic
tests/benchmarks/test_benchmarks.py0%
540.5 µs540 µs
test_get_model_cost_key_exact_match
tests/benchmarks/test_benchmarks.py0%
79.7 µs79.6 µs
test_get_model_cost_key_case_insensitive
tests/benchmarks/test_benchmarks.py0%
82.1 µs82.1 µs
test_get_llm_provider_openai
tests/benchmarks/test_benchmarks.py0%
137.7 µs138 µs
test_get_llm_provider_anthropic
tests/benchmarks/test_benchmarks.py0%
141.2 µs141.5 µs
test_token_counter_multi_turn
tests/benchmarks/test_benchmarks.py0%
565.7 µs567.2 µs
Commits
Click on a commit to change the comparison rangeBase
main
72a461b
+0.38%
fix(streaming): word-sliced cache replay for stream=true cache hits
199c7a3
16 days ago
by michelligabriele