issues
search
tensorzero
/
tensorzero
TensorZero creates a feedback loop for optimizing LLM applications — turning production data into smarter, faster, and cheaper models.
https://tensorzero.com
Apache License 2.0
656
stars
27
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Inference params in Inference and Batch Inference handlers are heavy and cloned at the start of the variant
#479
virajmehta
opened
30 minutes ago
1
Users should be able to limit the number of failovers per-function
#478
virajmehta
opened
3 hours ago
0
Configurable credentials
#477
virajmehta
opened
23 hours ago
1
Generate `inference_id` when the request begins and include in relevant logs
#476
GabrielBianconi
opened
1 day ago
0
Explore how to track partial failed streaming inferences
#475
GabrielBianconi
opened
1 day ago
2
Gateway doesn't catch response codes for streaming errors but just uses server
#474
virajmehta
opened
1 day ago
0
added instrumentation & tracing for top-level handlers, enforced nice…
#473
virajmehta
closed
1 day ago
0
bumped aiohttp version
#472
virajmehta
closed
3 days ago
0
If input validation fails in `/inference`, we should include the function name in the error message
#471
virajmehta
closed
1 day ago
0
Bump `aiohttp` version in example
#470
GabrielBianconi
closed
3 days ago
0
Bump version
#469
virajmehta
closed
3 days ago
0
fixed bug in client and wrote tests
#468
virajmehta
closed
3 days ago
0
Make Minijinja error message more useful when a variable is missing
#467
GabrielBianconi
closed
1 day ago
1
Support key-value tags in inference requests
#465
virajmehta
closed
3 days ago
0
bumped version
#464
virajmehta
closed
1 week ago
0
Support tagging inferences the way we tag feedbacks
#463
virajmehta
closed
3 days ago
0
Change default startup behavior to not require ClickHouse
#462
virajmehta
closed
1 week ago
0
TensorZero gateway should not fail on startup if the ClickHouse connection is missing
#461
virajmehta
closed
1 week ago
0
Update CONTRIBUTING.md
#460
GabrielBianconi
closed
1 week ago
0
added instrumentation to annotate error logging for model failures and such
#459
virajmehta
closed
3 days ago
0
Add `CONTRIBUTING.md`
#458
GabrielBianconi
closed
1 week ago
0
Add an integration for the xAI API
#457
virajmehta
opened
1 week ago
0
Add an integration for SGLang
#456
virajmehta
opened
1 week ago
0
bumped version
#455
virajmehta
closed
1 week ago
0
Add examples for the provider guides
#454
GabrielBianconi
closed
1 week ago
0
Add a "debug mode" for the gateway where we actually show request / response data in errors
#453
virajmehta
opened
1 week ago
0
Errors from GCP Vertex AI provider are impossible to debug
#452
GabrielBianconi
opened
1 week ago
0
Proposal: Implement `anyhow` for our errors
#451
virajmehta
closed
2 days ago
1
Dynamic credentials should be specific to an individual provider
#450
GabrielBianconi
opened
1 week ago
0
Test non-OpenAI models on Azure
#449
GabrielBianconi
opened
1 week ago
0
Gateway only supports (and requires) one credential per provider
#448
GabrielBianconi
opened
1 week ago
1
Write a RELEASE_GUIDE.md so we don't forget anything in a release
#447
virajmehta
opened
1 week ago
0
added version to status handler
#446
virajmehta
closed
1 week ago
0
Update example performance reference
#445
emmanuel-ferdman
closed
1 week ago
0
Implement an OpenAI-compatible TensorZero endpoint
#444
virajmehta
closed
1 week ago
0
Update copy
#443
GabrielBianconi
closed
1 week ago
0
Add SFT form to dashboard (unwired)
#442
GabrielBianconi
opened
1 week ago
0
Added tags to Python client's feedback method
#441
virajmehta
closed
1 week ago
0
OpenAI fine tuning for boolean metrics
#440
virajmehta
opened
1 week ago
0
Automatically keep track of % of inferences that match output schema in JSON functions
#439
GabrielBianconi
opened
2 weeks ago
2
Fix assorted issues in Python
#438
virajmehta
closed
2 weeks ago
0
Error message from Python client is unhelpful when schema validation fails on inputs
#437
virajmehta
closed
2 weeks ago
0
JSON Schema parsing errors are unhelpful
#436
virajmehta
closed
2 weeks ago
5
Python client's `_prepare_inference_request` causes side-effects
#434
GabrielBianconi
closed
2 weeks ago
1
Fix feedback return type in TensorZero Python client
#433
GabrielBianconi
closed
2 weeks ago
0
Publish the content files behind the T0 docs
#432
GabrielBianconi
opened
2 weeks ago
0
Set up `CONTRIBUTING.md`
#431
GabrielBianconi
closed
1 week ago
0
Expose the functionality for config parsing and validation of the gateway via a lib.rs interface that uses wasm
#430
virajmehta
opened
2 weeks ago
0
Allow text blocks alongside tool calls in provider tests
#429
virajmehta
closed
2 weeks ago
0
Tests are failing more than is needed when models return text + tool call
#428
virajmehta
closed
2 weeks ago
0
Next