fix(deepinfra): optimize rerank by removing query duplication
#24851Comparing
ovuruska:fix/deepinfra-rerank-query-dedup
(891718d
) withmain
(08be1e5
)0%
- Untouched
- 16
Benchmarks
16 total
test_get_model_info_with_provider
tests/benchmarks/test_benchmarks.py+3%
88.2 µs85.6 µs
test_get_llm_provider_openai
tests/benchmarks/test_benchmarks.py+1%
144.3 µs142.5 µs
test_get_llm_provider_anthropic
tests/benchmarks/test_benchmarks.py+1%
147.7 µs146.2 µs
test_get_model_cost_key_case_insensitive
tests/benchmarks/test_benchmarks.py+1%
88.4 µs87.5 µs
test_cost_per_token_openai
tests/benchmarks/test_benchmarks.py+1%
546.5 µs541.4 µs
test_get_model_cost_key_exact_match
tests/benchmarks/test_benchmarks.py+1%
86.1 µs85.4 µs
test_cost_per_token_anthropic
tests/benchmarks/test_benchmarks.py+1%
546 µs542.3 µs
test_token_counter_simple_message
tests/benchmarks/test_benchmarks.py+1%
240.3 µs238.9 µs
test_token_counter_raw_text
tests/benchmarks/test_benchmarks.py0%
188 µs187.8 µs
test_get_model_info_openai
tests/benchmarks/test_benchmarks.py0%
88 µs87.9 µs
test_token_counter_with_tools
tests/benchmarks/test_benchmarks.py0%
417.7 µs417.4 µs
test_token_counter_long_content
tests/benchmarks/test_benchmarks.py0%
1.7 ms1.7 ms
test_get_llm_provider_azure
tests/benchmarks/test_benchmarks.py-1%
150.2 µs151 µs
test_token_counter_multi_turn
tests/benchmarks/test_benchmarks.py-1%
568.3 µs571.3 µs
test_get_model_info_anthropic
tests/benchmarks/test_benchmarks.py-1%
84.7 µs85.3 µs
test_get_llm_provider_with_prefix
tests/benchmarks/test_benchmarks.py-1%
143 µs144.2 µs
Commits
Click on a commit to change the comparison rangeBase
main
08be1e5
+0.43%
fix(deepinfra): use single-element queries array for rerank instead of duplicating per document
9d60251
8 hours ago
by ovuruska-0.23%
test: update deepinfra rerank tests to expect single-element queries array
18f7524
8 hours ago
by ovuruska+0.23%
style: apply black formatting
891718d
7 hours ago
by ovuruska