Skip to content

Actions: FlashML-org/FreeToken

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
54 workflow runs
54 workflow runs

Filter by Workflow

⌥ + click/return to exclude

Filter by Event

⌥ + click/return to exclude

Filter by Status

⌥ + click/return to exclude

Filter by Branch

⌥ + click/return to exclude

Filter by Actor

⌥ + click/return to exclude
Nightly wheels Nightly wheels #30: Scheduled
7s main
main
7s
[Bug] Checkpoint conversion fails for Qwen3.6-35B FP8: Tensor size mismatch in dequant_fp8_weight Label issues #14: Issue #402 opened by quanrennsxsb
7s
7s
--moe-cache-auto leaves no room for prefill: boots and serves decode, then the first 8k prefill OOMs and kills the worker (NVFP4 dense) Label issues #13: Issue #401 opened by salekseev
7s
7s
Help/pr 132 Unit tests (NVIDIA) #2: Pull request #400 opened by samuelishida
Action required samuelishida:help/pr-132
samuelishida:help/pr-132
Action required
Running Copilot Code Review Copilot #5: by Copilot AI
4m 1s main
main
4m 1s
5090+256G内存无法运行红帽的GLM5.3FlashNVFP4 Label issues #12: Issue #397 opened by Zanry
43s
43s
[Feature] int4 routed experts: AWQ -> Q4_1 and compressed-tensors pack-quantized -> Q4_0, with no new CUDA kernel (offering a PR) Label issues #11: Issue #396 opened by salekseev
9s
9s
[Bug] --max-output-tokens is ignored by /v1/chat/completions and /v1/messages (only /v1/responses honours it) Label issues #10: Issue #395 opened by salekseev
8s
8s
Support official nvidia/Qwen3.8-Flash-Next-NVFP4 model Label issues #9: Issue #394 opened by lonestriker
7s
7s
Nightly wheels Nightly wheels #29: Scheduled
8s main
main
8s
FreeToken-Intel: 35B MoE, live tokens off an Arc Pro B70 (SYCL / XPU port) Label issues #8: Issue #299 edited by aangelinsf
8s
8s
Desktop engine installer ignores existing FreeToken engine path and always installs a new venv in AppData Label issues #7: Issue #387 opened by MSM-App
9s
9s
Running Copilot Code Review Copilot #4: by Copilot AI
2m 40s main
main
2m 40s
Running Copilot Code Review Copilot #3: by Copilot AI
5m 5s main
main
5m 5s
Build on Windows? Label issues #6: Issue #384 opened by XeonG
9s
9s
--moe-cache-auto ignores --num-tokens / --num-pages: expert fill overruns the budget and the KV pool OOMs at boot Label issues #5: Issue #383 opened by gdevenyi
7s
7s
Move the The Cache config element in the console window for QOL improvement. Label issues #4: Issue #382 opened by bmgjet
9s
9s
Nightly wheels Nightly wheels #28: Scheduled
3m 47s main
main
3m 47s
[Bug]: qwen3_5_moe fails to convert/load compressed-tensors NVFP4 checkpoints with model.language_model prefix Label issues #3: Issue #381 opened by silvertakana
8s
8s
ci(container): настроить сборку образа Build container image #1: Pull request #380 opened by Fgeeha
Action required Fgeeha:ci/docker-images
Fgeeha:ci/docker-images
Action required
ci(container): publish official Docker images to GHCR Label issues #2: Issue #379 opened by Fgeeha
7s
7s
could not encode request: Unexpected message role. Label issues #1: Issue #376 opened by MCJEModder2026
7s
7s
Nightly wheels Nightly wheels #27: Scheduled
3m 43s main
main
3m 43s
Nightly wheels Nightly wheels #26: Scheduled
3m 14s main
main
3m 14s
feat(rocm): AMD ROCm runtime for RDNA3/4 with native Qwen3.5-MoE GGUF decode parity Unit tests (NVIDIA) #1: Pull request #217 synchronize by samuelishida
Action required samuelishida:feat/amd-rocm-gfx1100-support
samuelishida:feat/amd-rocm-gfx1100-support
Action required