Skip to content

Actions: lablup/mlxcel

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
135 workflow runs
135 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
fix(cli): validate boolean cache flags, fix CLI-over-env precedence (… CI #118: Commit a074dfb pushed by inureyes
38s main
main
38s
fix(cli): validate boolean cache flags, fix CLI-over-env precedence CI #117: Pull request #70 opened by inureyes
41s fix/cli-flag-validation-help-cleanup
fix/cli-flag-validation-help-cleanup
41s
fix: align parallel context sizing with slots (#69) CI #116: Commit 31a03ca pushed by inureyes
38s main
main
38s
fix: align parallel context sizing with slots CI #115: Pull request #69 opened by inureyes
40s issue-57-parallel-context-semantics
issue-57-parallel-context-semantics
40s
fix: tighten memory estimator preflight coverage CI #114: Commit d2e7a84 pushed by inureyes
48s main
main
48s
fix: tighten memory estimator preflight coverage CI #113: Pull request #68 opened by inureyes
46s issue-52-memory-estimator-audit
issue-52-memory-estimator-audit
46s
feat: unified memory estimator with mlxcel inspect and preflight (#67) CI #112: Commit 080fb3c pushed by inureyes
36s main
main
36s
feat: unified memory estimator with mlxcel inspect and generate/serve preflight CI #111: Pull request #67 opened by inureyes
1m 12s feature/issue-56-unified-memory-estimator
feature/issue-56-unified-memory-estimator
1m 12s
feat(hardware): add KV cache memory estimator with 256-token rounding… CI #110: Commit a436b97 pushed by inureyes
39s main
main
39s
feat(hardware): add KV cache memory estimator with 256-token rounding CI #109: Pull request #65 synchronize by inureyes
43s feature/issue-54-kv-cache-estimator
feature/issue-54-kv-cache-estimator
43s
feat(weights): add exact weight footprint from safetensors header (#64) CI #108: Commit 3b1e2b3 pushed by inureyes
44s main
main
44s
feat(hardware): add KV cache memory estimator with 256-token rounding CI #107: Pull request #65 synchronize by inureyes
39s feature/issue-54-kv-cache-estimator
feature/issue-54-kv-cache-estimator
39s
feat(weights): add exact weight footprint from safetensors header CI #106: Pull request #64 synchronize by inureyes
47s feature/issue-53-exact-weight-footprint
feature/issue-53-exact-weight-footprint
47s
CI CI #105: by inureyes
36s main
main
36s
feat(ffi): wrap MLX runtime memory APIs (active/peak/limit) CI #104: Pull request #66 opened by inureyes
40s feature/issue-55-mlx-memory-ffi
feature/issue-55-mlx-memory-ffi
40s
feat(hardware): add KV cache memory estimator with 256-token rounding CI #103: Pull request #65 opened by inureyes
34s feature/issue-54-kv-cache-estimator
feature/issue-54-kv-cache-estimator
34s
feat(weights): add exact weight footprint from safetensors header CI #102: Pull request #64 opened by inureyes
39s feature/issue-53-exact-weight-footprint
feature/issue-53-exact-weight-footprint
39s
fix(prompt_cache): make radix trie iterative to prevent stack overflo… CI #101: Commit f6c37ac pushed by inureyes
50s main
main
50s
fix(prompt_cache): make radix trie iterative to prevent stack overflow CI #100: Pull request #63 opened by inureyes
36s fix/prompt-cache-trie-iterative
fix/prompt-cache-trie-iterative
36s
perf(gemma3n): improve M5 decode bandwidth with pretransposed weights… CI #99: Commit 4434b23 pushed by inureyes
54s main
main
54s
perf(gemma3n): improve M5 decode bandwidth with pretransposed weights CI #98: Pull request #62 opened by inureyes
39s port/gemma3n-m5-decode-bandwidth
port/gemma3n-m5-decode-bandwidth
39s
fix(gemma3n): gate bf16 fused decode path off M5 Neural Accelerator (… CI #97: Commit 206a2d5 pushed by inureyes
42s main
main
42s
fix(gemma3n): gate bf16 fused decode path off M5 Neural Accelerator CI #96: Pull request #61 opened by inureyes
36s port/gemma3n-fused-decode-na-gate
port/gemma3n-fused-decode-na-gate
36s
perf(gemma3n): reduce bf16 decode AltUp/MLP graph overhead (#60) CI #95: Commit 0e4faff pushed by inureyes
38s main
main
38s
perf(gemma3n): reduce bf16 decode AltUp/MLP graph overhead CI #94: Pull request #60 opened by inureyes
38s port/gemma3n-bf16-decode-overhead
port/gemma3n-bf16-decode-overhead
38s