apunkt/plano
1
0
Fork 0
mirror of https://github.com/katanemo/plano.git synced 2026-06-17 15:25:17 +02:00
Code Issues Projects Releases Packages Wiki Activity

improve move

Browse source
This commit is contained in:
Adil Hafeez 2024-11-18 15:46:57 -08:00
parent 4252c0f2f2
commit 658c3cb499
1 changed files with 0 additions and 39 deletions
Download patch file Download diff file Expand all files Collapse all files

39
arch/envoy.template.yaml
View file

@ -321,23 +321,6 @@ static_resources:
service_name: llm_gateway
random_sampling:
value: {{ arch_tracing.random_sampling }}
custom_tags:
- tag: user_prompt
metadata:
kind:
request: {}
metadata_key:
key: llm_filter
path:
- key: user_prompt
- tag: time_to_first_token
metadata:
kind:
request: {}
metadata_key:
key: llm_filter
path:
- key: time_to_first_token
{% endif %}
stat_prefix: arch_listener_http
codec_type: AUTO
@ -389,28 +372,6 @@ static_resources:
"@type": type.googleapis.com/envoy.extensions.compression.gzip.compressor.v3.Gzip
memory_level: 3
window_bits: 10
- name: envoy.filters.http.lua
typed_config:
'@type': type.googleapis.com/envoy.extensions.filters.http.lua.v3.Lua
default_source_code:
inline_string: |
function envoy_on_response(response_handle)
local user_message = response_handle:headers():get("x-user-message")
if user_message then
response_handle:logInfo("setting x-user-message")
response_handle:streamInfo():dynamicMetadata():set("llm_filter", "user_prompt", user_message)
response_handle:headers():remove("x-user-message")
end
local time_to_first_token = response_handle:headers():get("x-time-to-first-token")
if time_to_first_token then
response_handle:logInfo("setting x-time-to-first-token")
response_handle:streamInfo():dynamicMetadata():set("llm_filter", "time_to_first_token", time_to_first_token)
response_handle:headers():remove("x-time-to-first-token")
end
end
- name: envoy.filters.http.wasm
typed_config:
"@type": type.googleapis.com/udpa.type.v1.TypedStruct