exo/lm-lib
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: Update model profile and statistics tracking

Browse Source 
- Updated model profile to support new cost parameters: costPerCachedInToken and costPerReasoningOutToken.
- Enhanced AgentRequestStatistic to track cached input tokens and reasoning output tokens.
- Modified CompletionStatistics to include cached input tokens and reasoning output tokens.
- Updated AzureAiClient, CachedClient, and other clients to handle new statistics fields.
- Introduced Timings class to capture detailed timing metrics for requests.
- Adjusted Usage class to include input and output token details.
- Updated properties files to include new cost parameters for cached and reasoning tokens.
This commit is contained in:
extio-de
2025-11-30 18:19:01 +01:00
parent a6087a0613
commit b7c1e90dc7
19 changed files with 362 additions and 44 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
src/main/java/de/extio/lmlib/profile/ModelProfile.java
View File

@@ -14,7 +14,9 @@ public record ModelProfile(
String url,
String apiKey,
BigDecimal costPerInToken,
BigDecimal costPerCachedInToken,
BigDecimal costPerOutToken,
BigDecimal costPerReasoningOutToken,
String reasoningEffort,
String reasoningSummaryDetails,
String category) {

10
src/main/java/de/extio/lmlib/profile/ModelProfileService.java
View File

@@ -76,11 +76,19 @@ public class ModelProfileService {
.map(value -> new BigDecimal(value))
.orElse(BigDecimal.ZERO)
.divide(new BigDecimal(1000000.0));
final var costPerCachedInToken = Optional.ofNullable(resource.getProperty("cost1MCachedInTokens"))
.map(Object::toString)
.map(value -> new BigDecimal(value).divide(new BigDecimal(1000000.0)))
.orElse(costPerInToken);
final var costPerOutToken = Optional.ofNullable(resource.getProperty("cost1MOutTokens"))
.map(Object::toString)
.map(value -> new BigDecimal(value))
.orElse(BigDecimal.ZERO)
.divide(new BigDecimal(1000000.0));
final var costPerReasoningOutToken = Optional.ofNullable(resource.getProperty("cost1MReasoningOutTokens"))
.map(Object::toString)
.map(value -> new BigDecimal(value).divide(new BigDecimal(1000000.0)))
.orElse(BigDecimal.ZERO);
final var reasoningEffort = Optional.ofNullable(resource.getProperty("reasoningEffort"))
.map(Object::toString)
.orElse(null);
@@ -103,7 +111,7 @@ public class ModelProfileService {
.orElse(null);
}
return new ModelProfile(promptTemplate, tokenEncoding, maxTokens, maxContextLength, temperature, topP, modelProvider, modelNameCfg, url, apiKey, costPerInToken, costPerOutToken, reasoningEffort, reasoningSummaryDetails, category);
return new ModelProfile(promptTemplate, tokenEncoding, maxTokens, maxContextLength, temperature, topP, modelProvider, modelNameCfg, url, apiKey, costPerInToken, costPerCachedInToken, costPerOutToken, costPerReasoningOutToken, reasoningEffort, reasoningSummaryDetails, category);
}
catch (final IOException e) {
LOGGER.error("Error while reading model profile", e);