issues
search
gptscript-ai
/
knowledge
Knowledge for GPTScript
https://gptscript-ai.github.io/knowledge/
Apache License 2.0
29
stars
14
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
chore: drop mupdf build tag in favor of build constraint
#161
iwilltry42
closed
2 weeks ago
0
add: rolling markdown splitter
#160
iwilltry42
opened
2 weeks ago
0
feat: otto knowledge refactoring
#159
iwilltry42
closed
2 weeks ago
0
change: finish markdown_basic splitter and set it as default
#158
iwilltry42
opened
3 weeks ago
0
fix (openai embedding): retry on transient stream error when reading embeddings response
#157
iwilltry42
closed
3 weeks ago
0
change: simpler way of handling metadata inserted by .knowledge.json
#156
iwilltry42
closed
3 weeks ago
0
fix (postgres/breaking): change postgres document content columns to bytea to avoid encoding issue
#155
iwilltry42
closed
3 weeks ago
0
change: by default, do not exit on a failing file ingestion, but make sure we log the failed status
#154
iwilltry42
closed
3 weeks ago
0
change: use default concurrency of 100 for embeddings in pgvector
#153
iwilltry42
closed
3 weeks ago
0
fix: use pgxpool instead of pgxconn to avoid 'conn busy' errors
#152
iwilltry42
closed
3 weeks ago
0
add: structured json dataloader
#151
iwilltry42
closed
3 weeks ago
0
add: sqlite-vec vectorDB implementation
#150
iwilltry42
closed
3 weeks ago
0
Add delete file options, don't run prune
#149
StrongMonkey
closed
3 weeks ago
0
feat: structured ingestion
#148
iwilltry42
closed
2 weeks ago
1
feat (pgvector): enable HNSW by default
#147
iwilltry42
opened
4 weeks ago
0
feat: postgres index
#146
iwilltry42
closed
4 weeks ago
0
add: use .knowignore in top-level ingestion path if exists
#145
iwilltry42
closed
4 weeks ago
0
add: knowledge delete-file & knowledge get-file
#144
iwilltry42
closed
1 month ago
0
Unable to build using latest package
#143
njuptlzf
closed
2 weeks ago
1
fix: unique constraint bug in pgvector caused by duplicate creation of default dataset
#142
iwilltry42
closed
1 month ago
0
change (potentially breaking): use text-embedding-3-small by default everywhere (potentially breaking existing datasets, if mixed)
#141
iwilltry42
closed
1 month ago
0
add: pgvector vectorstore
#140
iwilltry42
closed
1 month ago
0
chore: new basic query tool used by otto
#139
ibuildthecloud
closed
1 month ago
0
fix: use credential keyword for credential references
#138
g-linville
closed
1 month ago
0
change: use model-provider tool instead of gateway credential
#137
iwilltry42
closed
1 month ago
0
fix: use a lock for extracting text from pdf via mupdf to avoid cgo panic (issue #135)
#136
iwilltry42
closed
1 month ago
0
bug: Recover from MuPDF panic
#135
iwilltry42
closed
4 weeks ago
0
Fix: use abs path to look for metadata
#134
StrongMonkey
closed
3 weeks ago
0
fix: make tool.gpt and gateway/tool.gpt match
#133
iwilltry42
closed
1 month ago
0
Fix: update to latest embedding model, apply timeout and corrupted collection fixes
#132
StrongMonkey
closed
1 month ago
0
Tweak more status from knowledge
#131
StrongMonkey
closed
1 month ago
0
Fix: add proper status to finished state and unsupported files
#130
StrongMonkey
closed
1 month ago
2
Fix: add embedding function back
#129
StrongMonkey
closed
1 month ago
0
Fix: Remove embedding logs to make output concise
#128
StrongMonkey
closed
1 month ago
0
feat: add progress indication for embedding generation process
#127
iwilltry42
closed
1 month ago
0
fix: use persistentpre child func to make DEBUG and KNOW_JSON env vars and flags work
#126
iwilltry42
closed
1 month ago
0
Fix: make script_id and thread_id optional
#125
StrongMonkey
closed
1 month ago
0
feat: ingestion - include metadata from .knowledge.json on dir level
#124
iwilltry42
closed
1 month ago
0
fix: embedding request retry 400 error due to missing request body
#123
iwilltry42
closed
2 months ago
0
feat(breaking): add structured json log + make filename mandatory for ingestion
#122
iwilltry42
closed
2 months ago
0
chore: use python3 instead of python3.12
#121
thedadams
closed
2 months ago
0
fix: properly cut out the pure JSON result from the knowledge tool logs
#120
iwilltry42
closed
2 months ago
0
File Syncing for Otto
#119
StrongMonkey
closed
2 weeks ago
1
feat: accept metadata files (JSON or k/v format) to add extra metadata to ingested documents
#118
iwilltry42
closed
1 month ago
0
fix: do not error out on first ingestion failure (continue with next file)
#117
iwilltry42
closed
2 months ago
1
feat: per-file ingestion progress using structured (json) logs
#116
iwilltry42
closed
1 month ago
0
feat: otto integration
#115
iwilltry42
closed
1 month ago
0
feat: add ingest and delete tools
#114
thedadams
closed
2 months ago
2
enhance: run retrievers within the merging retriever concurrently
#113
iwilltry42
closed
2 months ago
0
enhance: pure JSON retrieval output
#112
iwilltry42
closed
2 months ago
0
Next