Radoslav Gerganov 68ee98ae18 server : return HTTP 400 if prompt exceeds context length (#16486) ...

In streaming mode when prompt exceeds context length, the server returns
HTTP 200 status code with a JSON error in the body.  This is very
confusing and inconsistent with all other inference engines which return
HTTP 4xx error in this case.

This patch fixes this problem and makes the server return HTTP 400 in
such cases.

2025-10-10 16:11:07 +02:00

..

batched-bench

cmake : Do not install tools on iOS targets (#15903)

2025-09-16 09:54:44 +07:00

cvector-generator

cmake : Do not install tools on iOS targets (#15903)

2025-09-16 09:54:44 +07:00

export-lora

cmake : Do not install tools on iOS targets (#15903)

2025-09-16 09:54:44 +07:00

gguf-split

ci : use smaller model (#16168)

2025-09-22 09:11:39 +03:00

imatrix

cmake : Do not install tools on iOS targets (#15903)

2025-09-16 09:54:44 +07:00

llama-bench

llama : add --no-host to disable host buffers (#16310)

2025-10-06 19:55:53 +02:00

main

llama-cli: prevent spurious assistant token (#16202)

2025-09-29 10:03:12 +03:00

mtmd

chat : Granite Docling stopping (#16438)

2025-10-06 18:59:40 +02:00

perplexity

perplexity : show more kl-divergence data (#16321)

2025-09-29 09:30:45 +03:00

quantize

ci : use smaller model (#16168)

2025-09-22 09:11:39 +03:00

rpc

rpc : update documentation (#16441)

2025-10-07 06:59:13 +00:00

run

common: introduce http.h for httplib-based client (#16373)

2025-10-01 20:22:18 +03:00

server

server : return HTTP 400 if prompt exceeds context length (#16486)

2025-10-10 16:11:07 +02:00

tokenize

cmake : Do not install tools on iOS targets (#15903)

2025-09-16 09:54:44 +07:00

tts

model : Apertus model implementation (#15852)

2025-10-02 20:43:22 +03:00

CMakeLists.txt

mtmd : rename llava directory to mtmd (#13311)

2025-05-05 16:02:55 +02:00