issues
search
OpenRouterTeam
/
openrouter-runner
Inference engine powering open source models on OpenRouter
https://openrouter.ai
MIT License
512
stars
45
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
the first few sentences make sense and then it's just rubbish
#100
ErykCh
opened
2 days ago
3
Support for the parameter `n` in Requests
#99
ZachL1
opened
2 weeks ago
0
Support tools on DeepSeek Provider models
#98
jmfirth
opened
1 month ago
0
Logprobs not returning from OpenAI
#97
0xTomDaniel
closed
2 months ago
2
Incorrect response from "anthropic/claude-3.5-sonnet" model on OpenRouter API
#96
ghominejad
closed
2 months ago
0
wrong with shell AI website
#95
onlyhuman028
closed
3 months ago
2
deps: bump modal to 0.62.124
#94
sambarnes
closed
4 months ago
0
Feature Request: Support tools for Anthropic
#93
HavenDV
closed
3 months ago
2
fix: remove deprecated models bagel & psyfighter1
#92
sambarnes
closed
4 months ago
0
Does it support openai text embedding-3
#91
Harry-Lucas
opened
4 months ago
0
Can support dall-e-3 model ？
#90
gyp719
opened
4 months ago
0
Inquiry on Direct PDF Summarization Capability with OpenRouter
#89
KGZaker
closed
4 months ago
0
调用 gpt 模型 如何生成 图片？
#88
gyp719
closed
5 months ago
1
perf: let noromaid mixtral scale to zero
#87
sambarnes
closed
5 months ago
0
perf: let midnight rose scale to zero
#86
sambarnes
closed
5 months ago
0
perf: let bagel scale to zero
#85
sambarnes
closed
5 months ago
0
perf: reduce container idle timeout for neuralchat & psyfighter1
#84
sambarnes
closed
5 months ago
0
perf: always keep one midnight rose
#83
sambarnes
closed
5 months ago
0
chore: temporarily comment out keep_warm
#82
sambarnes
closed
5 months ago
0
perf: serve quantized Psyfighter2
#81
sambarnes
closed
5 months ago
0
perf: serve quantized versions of phi2, neuralchat, and psyfighter1
#80
sambarnes
closed
5 months ago
0
feat: add quantize_model() fn & a MidnightRose70B
#79
sambarnes
closed
5 months ago
0
deps: bump vllm from 0.2.6 to 0.3.3
#78
sambarnes
closed
2 months ago
0
perf: serve quantized versions of noromaid mixtral & bagel
#77
sambarnes
closed
6 months ago
0
fix: revert BACKLOG_THRESHOLD from 100 back to 30 following instability
#76
sambarnes
closed
6 months ago
0
feat: keep_warm=1 for noromaid mixtral & bagel
#75
sambarnes
closed
6 months ago
0
feat: add support for h100s & bump backlog limit
#74
sambarnes
closed
6 months ago
0
perf: bump noromaidmixtral to max_containers=3
#73
sambarnes
closed
6 months ago
0
fix: add a concurrency_limit to noromaidmixtral & bagel
#72
sambarnes
closed
6 months ago
0
refactor: move all models to their own unique containers
#71
sambarnes
closed
6 months ago
0
feat: add finish_reason to the protocol
#70
louisgv
closed
6 months ago
0
WIP: H100 support for Bagel
#69
alexanderatallah
closed
6 months ago
0
refactor: deprecate old usage fields & simplify vllm generate fn
#68
sambarnes
closed
6 months ago
0
docs: Update Docs to be more consistent.
#67
tyllenb
closed
6 months ago
1
Make model downloading synchronous again
#66
alexanderatallah
closed
7 months ago
0
Don't autoscale some containers
#65
alexanderatallah
closed
6 months ago
1
fix: running total of tokens for streams
#64
sambarnes
closed
7 months ago
1
fix: sentry scrubbing
#63
sambarnes
closed
7 months ago
1
feat: move phi-2 to a dedicated container, remove backup models
#62
louisgv
closed
7 months ago
0
feat: track completion request origin
#61
louisgv
closed
7 months ago
0
feat: timer estimates GPU costs
#60
sambarnes
closed
7 months ago
1
feat: remote generation on GPU timer to track costs
#59
sambarnes
closed
7 months ago
0
fix: key error when popping CUDA_VISIBLE_DEVICES env
#58
sambarnes
closed
7 months ago
0
fix: unset CUDA_VISIBLE_DEVICES in vLLM containers to patch Ray bug
#57
sambarnes
closed
7 months ago
1
perf: move large vLLM imports into the image.imports() context manager
#56
sambarnes
closed
6 months ago
0
ci: temporarily disable downloading models in CI, only deploy
#55
sambarnes
closed
7 months ago
1
fix: revert use of @enter entrypoint due to bug
#54
sambarnes
closed
8 months ago
0
perf: bump vllm container cpu memory from 128M to 1024M
#53
sambarnes
closed
7 months ago
0
ci: release to modal e2e env on pr label creation
#52
sambarnes
closed
2 months ago
0
ci: release to modal main env on push to main
#51
sambarnes
closed
8 months ago
0
Next