v0.0.24

What's Changed

• Properly handle completions in middle of document - Include LLM reque… by @jlewi in #271

Full Changelog: v0.0.23...v0.0.24

v0.0.23

What's Changed

• Evaluation - Simulate usage using examples produced from sessions by @jlewi in #253
• Render OpenAI request/response as HTML by @jlewi in #265
• Use gpt4-o-mini by default and filter out tags by @jlewi in #266
• Filter out blank responses by @jlewi in #270

Full Changelog: v0.0.22...v0.0.23

v0.0.22

What's Changed

• Sessionization and building an eval dataset by @jlewi in #230
• Fix Race Condition In Learning by @jlewi in #244
• Bug Fix: Learning doesn't happen because exampleDirs is empty by @jlewi in #245
• Reduce AI costs - Reduce context length by @jlewi in #243
• Enable RAG by default by @jlewi in #251
• Ensure embeddings matrix is never initialized with zero dimensions by @jlewi in #252
• Fix segfault in learner. by @jlewi in #256

Full Changelog: v0.0.21...v0.0.22

v0.0.21

What's Changed

• Learning Bug Fix: Need to enqueue blocks after execution events by @jlewi in #241

Full Changelog: v0.0.20...v0.0.21

v0.0.20

What's Changed

• LogAnalysis should not rely on cell execution being recorded by LogEvents by @jlewi in #222
• Analyzer should periodically checkpoint progress when processing large log files by @jlewi in #231
• Fix Learning: Cell Executions Not Being Added to BlockLogs by @jlewi in #233
• Fix troublsehooting guide so it shows up on the docs page. by @jlewi in #234
• Update the README to make the docs more prominent. by @jlewi in #235
• Enable RAG by default by @jlewi in #240
• Improve GetLLMLogs by @jlewi in #237

Full Changelog: v0.0.19...v0.0.20

v0.0.19

What's Changed

• Add doc about searching and extracting content using gmail. by @jlewi in #226
• Improve observability with OTEL so we can measure token usage and acceptance rate by @jlewi in #229

Full Changelog: v0.0.18...v0.0.19

v0.0.18

What's Changed

• Define an RPC to log events from the RunMe frontend. by @jlewi in #213
• Upgrade the default model to gpt4o by @jlewi in #218
• Change default HTTP port to 8877. by @jlewi in #219
• Troubleshooting guide for learning. by @jlewi in #216
• Replace getblocklog custom handler with a method defined in the LogsS… by @jlewi in #223
• Remove gRPC and gRPC gateway by @jlewi in #225

Full Changelog: v0.0.17...v0.0.18

v0.0.17

What's Changed

• Fix panic and error reporting in learner.Reconcile. by @jlewi in #201
• Add an RPC to get learned examples. by @jlewi in #202
• Define a Docker image for Foyle by @jlewi in #204
• Support GCP Logging by @jlewi in #207
• Don't hide the error when reconciling examples. by @jlewi in #209
• Allow Example Files To Be Overwritten by @jlewi in #210

Full Changelog: v0.0.16...v0.0.17

v0.0.16

What's Changed

• Generate Completions using Claude by @jlewi in #186
• Give each streaming generate request a unique trace id by @jlewi in #187
• Increase buffer size to fix bufio.Scanner: token too long by @jlewi in #189
• Log processing fixes for streaming by @jlewi in #190
• Add a traces service by @jlewi in #191
• Add GetLLMLogsRequest to look at actual LLM calls by @jlewi in #192
• Support loading markdown files for evaluation that don't have answers. by @jlewi in #193
• Update Foyle to filter exposition and return single code cell - Do well on single code cell generation by @jlewi in #194
• Update Runme Developer Docs. by @jlewi in #195
• Add a context id to streaming requests by @jlewi in #196
• Update the test. by @jlewi in #197
• Add an RPC to generate completions - Support FrontEnd using the connect protocol by @jlewi in #198

Full Changelog: v0.0.15...v0.0.16

v0.0.15

What's Changed

• Tech Note Auto Insert Suggested Cells by @jlewi in #161
• Update TN008 with details about GhostCells by @jlewi in #164
• Minor bugfixes by @jlewi in #165
• Server changes to support ghost cells and autocomplete by @jlewi in #166
• Publish protos to the BSR; upgrade buf to v2 config by @jlewi in #172
• Instrument the new AIService with OTEL connect so we generate trace ids. by @jlewi in #176
• Put technical notes on the website. by @jlewi in #177
• Level 1 Evals. by @jlewi in #181
• Prompt Tuning for Autocomplete by @jlewi in #184

Full Changelog: v0.0.14...v0.0.15