Skip to content

Actions: ml-explore/mlx-lm

Actions

Build and Test

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
pull_request.yml
711 workflow runs
711 workflow runs

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
Add optional fused SwitchGLU gate-up projection Build and Test #1479: Pull request #1319 synchronize by xxxkkw
Action required xxxkkw:switchglu-fused-gate-up
xxxkkw:switchglu-fused-gate-up
Action required
Fix qwen3_coder tool parser crash on non-numeric int/number params Build and Test #1478: Pull request #1330 opened by robertlangdonn
Action required robertlangdonn:fix-qwen3-coder-numeric-coercion
robertlangdonn:fix-qwen3-coder-numeric-coercion
Action required
Fix detokenizer for byte-level tokenizers with an SPM-style decoder (fixes #1041) Build and Test #1477: Pull request #1329 opened by robertlangdonn
Action required robertlangdonn:fix-tekken-detokenizer-spaces
robertlangdonn:fix-tekken-detokenizer-spaces
Action required
Fix Qwen3-Coder tool parser and harden server against mid-stream client disconnects Build and Test #1476: Pull request #1328 synchronize by i1rr
Action required i1rr:fix-qwen3-coder-nested-param-tags
i1rr:fix-qwen3-coder-nested-param-tags
Action required
Fix Qwen3-Coder tool parser and harden server against mid-stream client disconnects Build and Test #1475: Pull request #1328 opened by i1rr
Action required i1rr:fix-qwen3-coder-nested-param-tags
i1rr:fix-qwen3-coder-nested-param-tags
Action required
Fix mlx_lm.server 404 on short prompts (clamp negative start in think-token search) Build and Test #1474: Pull request #1327 opened by devYRPauli
Action required devYRPauli:fix/server-404-short-prompt
devYRPauli:fix/server-404-short-prompt
Action required
Add Step 3.7 Flash Build and Test #1473: Pull request #1325 opened by kernelpool
Action required kernelpool:add-step37-flash
kernelpool:add-step37-flash
Action required
Add BatchQuantizedKVCache Build and Test #1472: Pull request #1322 synchronize by xxxkkw
Action required xxxkkw:batch-quantized-kv-cache
xxxkkw:batch-quantized-kv-cache
Action required
Consolidate MLA kv_b_proj sanitize/shard into shared helpers Build and Test #1471: Pull request #1324 synchronize by scyyh11
Action required scyyh11:mla-sanitize-shard-helpers
scyyh11:mla-sanitize-shard-helpers
Action required
Consolidate MLA kv_b_proj sanitize/shard into shared helpers Build and Test #1470: Pull request #1324 opened by scyyh11
Action required scyyh11:mla-sanitize-shard-helpers
scyyh11:mla-sanitize-shard-helpers
Action required
Guard MLA caches from KV quantization Build and Test #1469: Pull request #1323 opened by xxxkkw
Action required xxxkkw:mla-cache-quant-guard
xxxkkw:mla-cache-quant-guard
Action required
Add BatchQuantizedKVCache Build and Test #1468: Pull request #1322 opened by xxxkkw
Action required xxxkkw:batch-quantized-kv-cache
xxxkkw:batch-quantized-kv-cache
Action required
Keep model parameters for fuse GGUF export Build and Test #1467: Pull request #1321 opened by abnormal749
Action required abnormal749:fix-fuse-gguf-donate-model
abnormal749:fix-fuse-gguf-donate-model
Action required
Fix Qwen3.5 MTP sanitize norm shift Build and Test #1466: Pull request #1320 opened by xxxkkw
Action required xxxkkw:qwen35-mtp-sanitize-norm
xxxkkw:qwen35-mtp-sanitize-norm
Action required
Add hybrid prefix cache restore coverage Build and Test #1465: Pull request #1314 synchronize by xxxkkw
Action required xxxkkw:hybrid-prefix-cache-contract
xxxkkw:hybrid-prefix-cache-contract
Action required
Add optional fused SwitchGLU gate-up projection Build and Test #1464: Pull request #1319 opened by xxxkkw
Action required xxxkkw:switchglu-fused-gate-up
xxxkkw:switchglu-fused-gate-up
Action required
Add opt-in recurrent profiling instrumentation Build and Test #1463: Pull request #1318 opened by xxxkkw
Action required xxxkkw:recurrent-profile-instrumentation
xxxkkw:recurrent-profile-instrumentation
Action required
Add olmo hybrid Build and Test #1460: Pull request #1315 synchronize by cmurray1105
Action required cmurray1105:add-olmo-hybrid
cmurray1105:add-olmo-hybrid
Action required
Add Prompt Lookup Decoding (ngram-simple) and Rolling-Hash Speculative Memory (ngram-mod) Build and Test #1459: Pull request #1297 synchronize by mayank2130
Action required mayank2130:pld-ngram-simple
mayank2130:pld-ngram-simple
Action required
Add olmo hybrid Build and Test #1458: Pull request #1315 synchronize by cmurray1105
Action required cmurray1105:add-olmo-hybrid
cmurray1105:add-olmo-hybrid
Action required
Add olmo hybrid Build and Test #1457: Pull request #1315 opened by cmurray1105
Action required cmurray1105:add-olmo-hybrid
cmurray1105:add-olmo-hybrid
Action required
Add hybrid prefix cache restore coverage Build and Test #1456: Pull request #1314 synchronize by xxxkkw
Action required xxxkkw:hybrid-prefix-cache-contract
xxxkkw:hybrid-prefix-cache-contract
Action required
Add hybrid prefix cache restore coverage Build and Test #1455: Pull request #1314 synchronize by xxxkkw
Action required xxxkkw:hybrid-prefix-cache-contract
xxxkkw:hybrid-prefix-cache-contract
Action required
Add hybrid prefix cache restore coverage Build and Test #1454: Pull request #1314 opened by xxxkkw
Action required xxxkkw:hybrid-prefix-cache-contract
xxxkkw:hybrid-prefix-cache-contract
Action required
Warn when speculative decoding may hurt throughput for MoE models Build and Test #1453: Pull request #1313 synchronize by Shylin26
Action required Shylin26:fix/moe-speculative-decoding-warning
Shylin26:fix/moe-speculative-decoding-warning
Action required