Skip to content

Actions: ml-explore/mlx-lm

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
792 workflow runs
792 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
Native MTP speculative decoding (Qwen3.5/3.6 reference implementation) Build and Test #1408: Pull request #990 synchronize by AirRunner
Action required AirRunner:feat/mtp-native
AirRunner:feat/mtp-native
Action required
Restrict think-state scan to assistant prefill tail Build and Test #1407: Pull request #1277 opened by eilidhmae
Action required eilidhmae:fix-think-state-user-content
eilidhmae:fix-think-state-user-content
Action required
Native MTP speculative decoding (Qwen3.5/3.6 reference implementation) Build and Test #1406: Pull request #990 synchronize by AirRunner
Action required AirRunner:feat/mtp-native
AirRunner:feat/mtp-native
Action required
Add TurboQuant KV cache + DeepSeek V4 support Build and Test #1405: Pull request #1067 synchronize by arozanov
Action required arozanov:feature/turboquant-kv-cache
arozanov:feature/turboquant-kv-cache
Action required
Add Gemma 4 assistant (MTP drafter) model class Build and Test #1404: Pull request #1276 opened by broomva
Action required broomva:feat/gemma4-assistant
broomva:feat/gemma4-assistant
Action required
fix: make generation_stream per-thread to fix server crash on worker threads Build and Test #1403: Pull request #1275 synchronize by nish2292
Action required nish2292:feat/fix-sliding-window-server-crash
nish2292:feat/fix-sliding-window-server-crash
Action required
fix: make generation_stream per-thread to fix server crash on worker threads Build and Test #1402: Pull request #1275 opened by nish2292
Action required nish2292:feat/fix-sliding-window-server-crash
nish2292:feat/fix-sliding-window-server-crash
Action required
feat: add --idle-timeout to unload model after inactivity Build and Test #1401: Pull request #1274 opened by nish2292
Action required nish2292:feat/idle-model-unloading
nish2292:feat/idle-model-unloading
Action required
Add TurboQuant KV cache + DeepSeek V4 support Build and Test #1400: Pull request #1067 synchronize by arozanov
Action required arozanov:feature/turboquant-kv-cache
arozanov:feature/turboquant-kv-cache
Action required
Add logits processor arguments to mlx_lm.generate Build and Test #1399: Pull request #1273 opened by realyxl
Action required realyxl:cli-logits-processors
realyxl:cli-logits-processors
Action required
Support max_kv_size configuration in HTTP server Build and Test #1398: Pull request #1272 opened by r-bahuguna
Action required r-bahuguna:feature/server-max-kv-size
r-bahuguna:feature/server-max-kv-size
Action required
Add TurboQuant KV cache + DeepSeek V4 support Build and Test #1397: Pull request #1067 synchronize by arozanov
Action required arozanov:feature/turboquant-kv-cache
arozanov:feature/turboquant-kv-cache
Action required
Add Responses API support Build and Test #1396: Pull request #1263 synchronize by blairhudson
Action required blairhudson:responses
blairhudson:responses
Action required
Add Responses API support Build and Test #1395: Pull request #1263 synchronize by blairhudson
Action required blairhudson:responses
blairhudson:responses
Action required
Fix LFM2.5 tool parser inference Build and Test #1394: Pull request #1260 synchronize by blairhudson
Action required blairhudson:fix/lfm25-tool-parser-inference
blairhudson:fix/lfm25-tool-parser-inference
Action required
Add Olmo3 tool parser Build and Test #1393: Pull request #1271 opened by anthonyhchan
Action required anthonyhchan:olmo3-tool-call
anthonyhchan:olmo3-tool-call
Action required
Native MTP speculative decoding (Qwen3.5/3.6 reference implementation) Build and Test #1392: Pull request #990 synchronize by AirRunner
Action required AirRunner:feat/mtp-native
AirRunner:feat/mtp-native
Action required
docs: add Qwen3 QLoRA Apple Silicon example Build and Test #1391: Pull request #1270 opened by SysCd
Action required SysCd:docs/qwen3-lora-clean
SysCd:docs/qwen3-lora-clean
Action required
feat: Drop unsupported weights by default during model load Build and Test #1389: Pull request #1033 synchronize by spicyneuron
Action required spicyneuron:drop-unknown-weights
spicyneuron:drop-unknown-weights
Action required
add Qwen3 QLoRA Apple Silicon example Build and Test #1388: Pull request #1269 opened by SysCd
Action required SysCd:docs/qwen3-lora-apple-silicon
SysCd:docs/qwen3-lora-apple-silicon
Action required
fix(server): wire --prompt-cache-bytes CLI flag to LRUPromptCache Build and Test #1387: Pull request #1267 opened by andreinknv
Action required andreinknv:fix/wire-prompt-cache-bytes-cli
andreinknv:fix/wire-prompt-cache-bytes-cli
Action required
Add TurboQuant KV cache + DeepSeek V4 support Build and Test #1386: Pull request #1067 synchronize by arozanov
Action required arozanov:feature/turboquant-kv-cache
arozanov:feature/turboquant-kv-cache
Action required
Add TurboQuant KV cache + DeepSeek V4 support Build and Test #1385: Pull request #1067 synchronize by arozanov
Action required arozanov:feature/turboquant-kv-cache
arozanov:feature/turboquant-kv-cache
Action required
feat(server): add /v1/embeddings route via mlx_embeddings Build and Test #1384: Pull request #1265 opened by andreinknv
Action required andreinknv:feat/embeddings-route
andreinknv:feat/embeddings-route
Action required
Native MTP speculative decoding (Qwen3.5/3.6 reference implementation) Build and Test #1383: Pull request #990 synchronize by AirRunner
Action required AirRunner:feat/mtp-native
AirRunner:feat/mtp-native
Action required