Commit History

Revert reasoning default to deepseek — Qwen3-4B-Thinking timed out live-testing at 90s
b478f20
Running

Jolly6443 commited on

Add newer/smaller default models per role: Qwen3-4B-Thinking (reasoning), Dolphin3.0-3B (uncensored)
f3aafd7

Jolly6443 commited on

Swap Dolphin 8B for Dolphin 7B (8B was too slow for free CPU tier, regularly hit the 90s self-heal timeout)
558d2c2
verified

Jolly6443 commited on

Sync: fix Qwen general-role model repo in shared catalogue
770e9e8
verified

Jolly6443 commited on

Fix: remove hardcoded ENV MODEL_ID=qwen that was silently overriding SPACE_ROLE on every Space
fd8d54e
verified

Jolly6443 commited on

Remove temp diagnostic endpoint (root cause found)
d5498d8
verified

Jolly6443 commited on

Temp diagnostic: add /jolly/debug-env
297eb68
verified

Jolly6443 commited on

Fix: queue concurrent requests instead of instant 429 rejection
b9dbe38
verified

Jolly6443 commited on

Document SPACE_ROLE multi-space setup and /jolly/refresh-model
911e357
verified

Jolly6443 commited on

Add role-based auto model assignment (SPACE_ROLE=reasoning) + /jolly/refresh-model
2fd12a2
verified

Jolly6443 commited on

Add 90s generation timeout that kills+restarts the worker instead of holding the busy-lock forever on an abandoned/oversized request
b8434c4

JollyDay Assistant commited on

Reduce max_tokens cap to 80 (~15s) — 180 tokens was still hitting a client-side timeout
9b9dc97

JollyDay Assistant commited on

Fail fast with 429 when already generating, instead of silently queuing behind it
58e8bc5

JollyDay Assistant commited on

Run inference in a dedicated child process; tighten max_tokens cap to 180 for tolerable response times without streaming
9c4a2b6

JollyDay Assistant commited on

Cap max_tokens at 400 — long CPU-only generations block the entire process, not just the request
e2184e8

JollyDay Assistant commited on

Rewrite: load model directly with Llama class, drop buggy llama_cpp.server subprocess/proxy entirely
2f079cf

JollyDay Assistant commited on

Fix empty-body proxy responses: reuse a long-lived httpx client instead of closing it before streaming finishes
37cb455

JollyDay Assistant commited on

Strip hop-by-hop headers in proxy forward — fixes empty 400s on all /v1/* routes
591b4ad

JollyDay Assistant commited on

Fix health check: llama_cpp.server has no /health route, poll /v1/models instead
4d114a9

JollyDay Assistant commited on

Switch base image to Alpine — the pinned wheel is musl-linked, not glibc
bc610c2

JollyDay Assistant commited on

Pin llama-cpp-python to fix build timeout (Job timeout / BUILD_ERROR)
1f1ea05

JollyDay Assistant commited on

Update app.py
366fe91
verified

Jolly6443 commited on

Update Dockerfile
da89010
verified

Jolly6443 commited on

Update app.py
efc88b9
verified

Jolly6443 commited on

Update app.py
a497fb4
verified

Jolly6443 commited on

Update Dockerfile
7c691d7
verified

Jolly6443 commited on

Update app.py
19c8916
verified

Jolly6443 commited on

Update Dockerfile
a776e34
verified

Jolly6443 commited on

Update README.md
cec20de
verified

Jolly6443 commited on

Update app.py
19180a4
verified

Jolly6443 commited on

Update Dockerfile
e06bd08
verified

Jolly6443 commited on

Update README.md
cf05585
verified

Jolly6443 commited on

Update app.py
e112c14
verified

Jolly6443 commited on

Update Dockerfile
f95c194
verified

Jolly6443 commited on

Update README.md
eaca05c
verified

Jolly6443 commited on

Create app.py
40d6ecf
verified

Jolly6443 commited on

Create Dockerfile
9e27039
verified

Jolly6443 commited on

initial commit
aafe80d
verified

Jolly6443 commited on