From f51c12aa871fa661159d8b0c8d61f112e68eec78 Mon Sep 17 00:00:00 2001 From: jiangkuaixue123 Date: Thu, 22 Oct 2026 19:56:51 +0800 Subject: [PATCH 1/8] docs(casebook): add three dedicated-engine cases from a Qwen3.8 campaign - CUDA Graph replay read capture-time state: greedy ids matched while per-step logits diverged by 0.70, so token-id equality is not a correctness oracle. - Silent capacity and error failures: HTTP 200 with an SSE error frame, HTTP 400 with an empty body and no log line, and a bucket-headroom failure at token 993. - Upstream FP8 kernels vs a self-written Triton GEMM: 2.8 vs 273 TFLOPS on the target card, with the pre-registered go/abandon thresholds and the two installation traps (placeholder PyPI package, vendored extension pinned to a torch minor version). Each entry records the observation, the disposition, a suggested revisit check, its applicability and its limitations; none makes a performance claim. Signed-off-by: jiangkuaixue123 Co-authored-by: jiangkuaixue123 --- .../inference-optimization-casebook/SKILL.md | 6 ++++ ...n38-cuda-graph-state-not-accumulating.json | 26 ++++++++++++++ ...wen38-cuda-graph-state-not-accumulating.md | 34 ++++++++++++++++++ ...38-silent-capacity-and-error-failures.json | 25 +++++++++++++ ...en38-silent-capacity-and-error-failures.md | 28 +++++++++++++++ ...38-upstream-fp8-kernels-vs-own-triton.json | 26 ++++++++++++++ ...en38-upstream-fp8-kernels-vs-own-triton.md | 36 +++++++++++++++++++ 7 files changed, 181 insertions(+) create mode 100644 src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.json create mode 100644 src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.md create mode 100644 src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.json create mode 100644 src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md create mode 100644 src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.json create mode 100644 src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.md diff --git a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/SKILL.md b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/SKILL.md index abf4d37d1..851c977f9 100644 --- a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/SKILL.md +++ b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/SKILL.md @@ -86,6 +86,12 @@ Current owner-local records retain rejected directions alongside their limits: attention/cache alternatives and missing gain on one 910B1 cohort. - [Drama3 B300](cases/drama3-rejected-directions.md): invalid controls, correct kernels without service gain, prefix reuse and analytical rejection. +- [CUDA Graph state not accumulating](cases/qwen38-cuda-graph-state-not-accumulating.md): + token ids matched while per-step logits diverged, so ids are not a correctness oracle. +- [Silent capacity and error failures](cases/qwen38-silent-capacity-and-error-failures.md): + HTTP 200 with an error frame, HTTP 400 with an empty body, failure at token 993. +- [Upstream FP8 kernels vs own Triton](cases/qwen38-upstream-fp8-kernels-vs-own-triton.md): + a numerically cautious kernel ran 100x slower than the borrowed library. Each `.json` sibling owns identity, origin, source pins and limitations; the Markdown owns the observation and suggested revisit check. Read them as diff --git a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.json b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.json new file mode 100644 index 000000000..1b800e203 --- /dev/null +++ b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.json @@ -0,0 +1,26 @@ +{ + "id": "case.qwen38-cuda-graph-state-not-accumulating", + "kind": "case", + "origin": ["first_party"], + "state": "historical_observation", + "content": "qwen38-cuda-graph-state-not-accumulating.md", + "sources": [ + { + "type": "local", + "note": "Qwen3.8 dedicated-engine campaign, run records for the CUDA Graph round", + "paths": ["evidence/"] + } + ], + "applicability": [ + "Single device, batch size 1", + "Hybrid linear-attention model (48 DeltaNet + 16 full-attention layers)", + "PyTorch CUDA Graph capture of one decode step" + ], + "tags": ["cuda-graph", "correctness-gate", "state-management", "silent-wrong-result"], + "limitations": [ + "One H20-3e card, one workload; no concurrency", + "Graph path acceptance recorded INCONCLUSIVE after the timing run", + "Does not establish behaviour for multi-request or paged-state graphs" + ], + "attribution": "Dedicated-engine campaign; defect and fix verified by per-step logits comparison" +} diff --git a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.md b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.md new file mode 100644 index 000000000..946f35ee1 --- /dev/null +++ b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.md @@ -0,0 +1,34 @@ +# CUDA Graph replay read capture-time state: ids stayed equal while logits diverged + +A single-request Qwen3.8 decode engine (64 layers: 48 linear-attention/DeltaNet +plus 16 full-attention) captured its whole decode step into a CUDA Graph. Greedy +token ids matched the eager path exactly, so three successive regression gates +passed. Per-step logits comparison later showed the graph path was not computing +the same thing: the hybrid state was rebuilt as fresh tensors every step, so +replay kept reading the addresses captured on the first step and never +accumulated state. + +| Observation | Value | +| --- | --- | +| Per-step max_abs logits, graph vs eager | 0.70 (greedy ids identical) | +| After preallocating and writing state in place | 0.00e+00 on the tiny fixture | +| TPOT delivered by the affected configuration | unchanged (the bug costs no time) | + +Disposition: a token-id equality gate is not a correctness oracle. A graph path +needs a per-step logits comparison against eager, plus an explicit statement of +which buffers are captured by address. The fix is to preallocate every recurrent +buffer the graph writes and to capture into a dedicated memory pool so later +eager allocations cannot alias it. + +Revisit condition: any engine that captures a step whose state is created inside +the step. First check: print the data pointers of every state tensor before and +after capture; if they change between steps, replay is writing to stale storage. + +Applicability: single device, batch size 1, hybrid linear-attention models, +PyTorch CUDA Graph. Not established for multi-request graphs or for graphs whose +state lives in a paged allocator. + +Limitations: observed on one H20-3e card with one workload; no concurrency and +no cross-device qualification. The configuration was recorded as INCONCLUSIVE +for acceptance purposes because the measurement that exposed it happened after +the timing run. diff --git a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.json b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.json new file mode 100644 index 000000000..927b67a65 --- /dev/null +++ b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.json @@ -0,0 +1,25 @@ +{ + "id": "case.qwen38-silent-capacity-and-error-failures", + "kind": "case", + "origin": ["first_party"], + "state": "historical_observation", + "content": "qwen38-silent-capacity-and-error-failures.md", + "sources": [ + { + "type": "local", + "note": "Qwen3.8 dedicated-engine campaign, serving-layer debug rounds", + "paths": ["evidence/"] + } + ], + "applicability": [ + "OpenAI-compatible HTTP wrapper over a custom inference engine", + "Bucketed CUDA Graph decode with a headroom-selected bucket" + ], + "tags": ["serving", "error-handling", "observability", "capacity", "silent-failure"], + "limitations": [ + "One engine, one card", + "Behaviour past capacity not characterised beyond the failure itself", + "No concurrency or multi-client qualification" + ], + "attribution": "Dedicated-engine campaign; each symptom reproduced and fixed" +} diff --git a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md new file mode 100644 index 000000000..bb5b17707 --- /dev/null +++ b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md @@ -0,0 +1,28 @@ +# Three ways an inference server reported success while failing + +While measuring a single-request engine through its OpenAI-compatible HTTP +endpoint, three distinct failures produced no actionable evidence: two returned +a success status, one returned an empty error body. Each is cheap to detect and +each cost a debugging round. + +| Symptom | What actually happened | Detection that would have caught it | +| --- | --- | --- | +| HTTP 200 with an SSE `error` frame mid-stream | Decode ran past the configured graph bucket, the engine fell back or failed, and the wrapper kept the stream open | Assert `completion_tokens` equals the requested count and `finish_reason == length` on every request | +| HTTP 400, empty body, 0-byte server log | The error handler deliberately dropped all exceptions ("client addresses and arbitrary exceptions do not belong in service logs") | Log a structured line with exception type and a truncated message; never return an error status without a body | +| Generation failed at token 993 with no message | The graph bucket was chosen from `prompt + headroom` with `headroom = 256`, while the request asked for 2048 output tokens | Validate bucket coverage against maximum output length, not against prompt slack, before starting the run | + +Disposition: a serving layer must never convert an internal failure into a +success response. Log the exception type to stderr even when the payload must be +redacted, return a JSON error body for every 4xx/5xx, and treat capacity +(bucket, context, output length) as a pre-flight assertion rather than a runtime +surprise. + +Revisit condition: any wrapper that owns stream framing or error translation. +First check: send one request whose output length exceeds the configured +capacity and confirm the client sees a non-2xx status with a body. + +Applicability: OpenAI-compatible serving wrappers over a custom engine, any +device. The third row is specific to bucketed graph execution. + +Limitations: observed on one engine and one card; the exact fallback behaviour +past capacity was not characterised beyond "generation failed". diff --git a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.json b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.json new file mode 100644 index 000000000..bf2a8a28b --- /dev/null +++ b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.json @@ -0,0 +1,26 @@ +{ + "id": "case.qwen38-upstream-fp8-kernels-vs-own-triton", + "kind": "case", + "origin": ["first_party"], + "state": "historical_observation", + "content": "qwen38-upstream-fp8-kernels-vs-own-triton.md", + "sources": [ + { + "type": "local", + "note": "Qwen3.8 dedicated-engine campaign, FP8 kernel microbenchmarks", + "paths": ["evidence/"] + } + ], + "applicability": [ + "SM90-class accelerator, FP8 e4m3 with 128x128 block scales", + "M in the thousands (prefill), batch size 1 decode measured separately", + "torch 2.14 engine environment" + ], + "tags": ["fp8", "gemm", "build-vs-borrow", "microbenchmark", "licensing"], + "limitations": [ + "Single card; numbers are microbenchmarks, not end-to-end", + "Thresholds registered for one shape family and one device", + "Upstream versions may change both performance and licensing" + ], + "attribution": "Dedicated-engine campaign; measurements taken on the target card" +} diff --git a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.md b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.md new file mode 100644 index 000000000..802d2f1c6 --- /dev/null +++ b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.md @@ -0,0 +1,36 @@ +# A self-written block-scaled FP8 GEMM was 100x slower than the upstream kernel + +A dedicated engine shipped a Triton block-scaled FP8 GEMM that padded 32 +singleton lanes into an M64 tile to keep numerics stable. It ran at 2.8 TFLOPS +on H20-3e and made an 8K prefill take 145 s. Measuring the upstream libraries on +the same shapes turned the build-versus-borrow decision into arithmetic. + +| Implementation | 8192x17408x5120 | Note | +| --- | --- | --- | +| Own Triton block-scaled kernel | 2.8 TFLOPS | ~1% of FP8 peak | +| FlashInfer block-scaled (installed from PyPI, Apache-2.0) | 273 TFLOPS | usable in the engine environment | +| DeepGEMM (vendored inside vLLM, prebuilt `.so`) | 279 TFLOPS | unusable outside vLLM's environment: the extension pins a torch minor version | + +Thresholds were registered before the measurement: go at >=280 TFLOPS, abandon +below 250. FlashInfer measured 273, so it was adopted with the 2.5% shortfall +recorded rather than rounded away. The same library reports 3273 GB/s for M=1, +which makes it a candidate for the bandwidth-bound decode path as well. + +Two traps worth recording. `pip install deep_gemm` resolves to an unrelated +18 KB placeholder on PyPI; the real project is source-only. And a vendored +extension compiled against one torch minor version fails at import on another, +so "reuse the dependency of the reference server" is not a shortcut. + +Disposition: measure the upstream implementation on your own shapes before +writing a kernel whose only claim is numerical caution. Register the go/abandon +threshold first; record the shortfall if you adopt below it. + +Revisit condition: new shapes, a different accelerator, or a new upstream +release. First check: rerun the shape sweep and confirm the largest shapes still +sit near peak before reusing this decision. + +Applicability: SM90-class device, FP8 e4m3 with 128x128 block scales, torch 2.14 +engine environment. The M=1 numbers are microbenchmarks, not end-to-end decode. + +Limitations: single-card microbenchmarks; end-to-end prefill was measured +separately; upstream versions may change both the numbers and the licensing. From 9317840d49f40762190072b7cc7d6d1848bd7179 Mon Sep 17 00:00:00 2001 From: jiangkuaixue123 Date: Sat, 31 Oct 2026 09:46:39 +0000 Subject: [PATCH 2/8] docs(casebook): bind Qwen3.8 cases to retained records Point each case at the issue ledger, engine blob, or retained evidence file that carries the observation, including the digest and the figures those records do not contain. Agent-assisted-by: Cursor / grok-4.7 [effort=unknown] (author) Signed-off-by: jiangkuaixue123 Co-authored-by: jiangkuaixue123 --- ...n38-cuda-graph-state-not-accumulating.json | 75 +++++++++++------ ...wen38-cuda-graph-state-not-accumulating.md | 9 +++ ...38-silent-capacity-and-error-failures.json | 80 +++++++++++++------ ...en38-silent-capacity-and-error-failures.md | 11 +++ ...38-upstream-fp8-kernels-vs-own-triton.json | 74 +++++++++++------ ...en38-upstream-fp8-kernels-vs-own-triton.md | 11 +++ 6 files changed, 189 insertions(+), 71 deletions(-) diff --git a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.json b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.json index 1b800e203..df7fdd05a 100644 --- a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.json +++ b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.json @@ -1,26 +1,53 @@ { - "id": "case.qwen38-cuda-graph-state-not-accumulating", - "kind": "case", - "origin": ["first_party"], - "state": "historical_observation", - "content": "qwen38-cuda-graph-state-not-accumulating.md", - "sources": [ - { - "type": "local", - "note": "Qwen3.8 dedicated-engine campaign, run records for the CUDA Graph round", - "paths": ["evidence/"] - } - ], - "applicability": [ - "Single device, batch size 1", - "Hybrid linear-attention model (48 DeltaNet + 16 full-attention layers)", - "PyTorch CUDA Graph capture of one decode step" - ], - "tags": ["cuda-graph", "correctness-gate", "state-management", "silent-wrong-result"], - "limitations": [ - "One H20-3e card, one workload; no concurrency", - "Graph path acceptance recorded INCONCLUSIVE after the timing run", - "Does not establish behaviour for multi-request or paged-state graphs" - ], - "attribution": "Dedicated-engine campaign; defect and fix verified by per-step logits comparison" + "id": "case.qwen38-cuda-graph-state-not-accumulating", + "kind": "case", + "origin": ["first_party"], + "state": "historical_observation", + "content": "qwen38-cuda-graph-state-not-accumulating.md", + "sources": [ + { + "type": "internal", + "issue": "https://github.com/InferMatrixLab/InferRSI/issues/67", + "run": "issuecomment-6337245823", + "artifact": "https://github.com/InferMatrixLab/InferRSI/issues/67#issuecomment-6337245823", + "version": "2026-10-22T09:18:18Z", + "sha256": "de68a42eb35593b850c63e9bdc29f6896775d8f15fa34fdc8f8be8a6866a7f3f", + "integrity_status": "issue_comment_body_sha256", + "note": "Issue #67 ledger: v11 state did not accumulate; greedy ids matched while logits differed by 0.70 (67 ULP). Digest is the comment body retrieved 2026-10-31." + }, + { + "type": "internal", + "issue": "https://github.com/InferMatrixLab/InferRSI/issues/67", + "run": "store:case.qwen38-graph-state-not-accumulating-cand", + "artifact": "https://github.com/InferMatrixLab/InferRSI/issues/67#issuecomment-6331682106", + "version": "2026-10-22T03:06:13Z", + "sha256": "9f7d7085e4d4508712fa5f5a1022111092ad897703663c697773e20bcccd0663", + "integrity_status": "issue_comment_body_sha256", + "note": "Names the store candidate and records a v11 tiny-fixture bitwise pass. The public prefix 5c7700d is described there as a whitelist checkout, not an object in the public engine repository. The comment body does not contain the literal 0.00e+00." + }, + { + "type": "git", + "url": "https://github.com/jiangkuaixue123/nano-vllm/blob/3f7516a7b4a2c0a34871a453c294bc23b1b6c67c/nanovllm/qwen38/graph.py", + "commit": "3f7516a7b4a2c0a34871a453c294bc23b1b6c67c", + "paths": ["nanovllm/qwen38/graph.py"], + "content_sha256": "d4213aa68599564a01ced3b990993c9b17482615c154120bc4180f78a5da6098", + "issue": "https://github.com/InferMatrixLab/InferRSI/issues/67", + "note": "Public engine tree named by the same ledger. Dedicated capture pool and in-place KV, convolution and recurrent buffers. This is the fix disposition, not the 0.70 measurement." + } + ], + "applicability": [ + "Single device, batch size 1", + "Hybrid linear-attention model (48 DeltaNet + 16 full-attention layers)", + "PyTorch CUDA Graph capture of one decode step" + ], + "tags": ["cuda-graph", "correctness-gate", "state-management", "silent-wrong-result"], + "limitations": [ + "One H20-3e card, one workload; no concurrency", + "Graph path acceptance recorded INCONCLUSIVE after the timing run", + "Does not establish behaviour for multi-request or paged-state graphs", + "The 0.70 figure is the Issue #67 comment body above; raw logits were not rehashed for this binding", + "The tiny-fixture pass is recorded as bitwise in that ledger; the literal 0.00e+00 is this case's notation for that pass", + "Prefix 5c7700d was not resolved to a public 40-character commit" + ], + "attribution": "Issue #67 ledger comments 6337245823 and 6331682106, plus nano-vllm graph.py at 3f7516a7b4a2c0a34871a453c294bc23b1b6c67c" } diff --git a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.md b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.md index 946f35ee1..5206f20c8 100644 --- a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.md +++ b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.md @@ -32,3 +32,12 @@ Limitations: observed on one H20-3e card with one workload; no concurrency and no cross-device qualification. The configuration was recorded as INCONCLUSIVE for acceptance purposes because the measurement that exposed it happened after the timing run. + +Retained records: the 0.70 gap is Issue #67 comment 6337245823 (body sha256 +`de68a42eb35593b850c63e9bdc29f6896775d8f15fa34fdc8f8be8a6866a7f3f`). Comment +6331682106 names `store:case.qwen38-graph-state-not-accumulating-cand` and a +tiny-fixture bitwise pass (body sha256 +`9f7d7085e4d4508712fa5f5a1022111092ad897703663c697773e20bcccd0663`); it does not +contain the literal `0.00e+00`, and prefix `5c7700d` is not a public commit. +The in-place buffers and dedicated capture pool are `nanovllm/qwen38/graph.py` +at `3f7516a7b4a2c0a34871a453c294bc23b1b6c67c`. diff --git a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.json b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.json index 927b67a65..ab29841e0 100644 --- a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.json +++ b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.json @@ -1,25 +1,59 @@ { - "id": "case.qwen38-silent-capacity-and-error-failures", - "kind": "case", - "origin": ["first_party"], - "state": "historical_observation", - "content": "qwen38-silent-capacity-and-error-failures.md", - "sources": [ - { - "type": "local", - "note": "Qwen3.8 dedicated-engine campaign, serving-layer debug rounds", - "paths": ["evidence/"] - } - ], - "applicability": [ - "OpenAI-compatible HTTP wrapper over a custom inference engine", - "Bucketed CUDA Graph decode with a headroom-selected bucket" - ], - "tags": ["serving", "error-handling", "observability", "capacity", "silent-failure"], - "limitations": [ - "One engine, one card", - "Behaviour past capacity not characterised beyond the failure itself", - "No concurrency or multi-client qualification" - ], - "attribution": "Dedicated-engine campaign; each symptom reproduced and fixed" + "id": "case.qwen38-silent-capacity-and-error-failures", + "kind": "case", + "origin": ["first_party"], + "state": "historical_observation", + "content": "qwen38-silent-capacity-and-error-failures.md", + "sources": [ + { + "type": "git", + "url": "https://github.com/jiangkuaixue123/nano-vllm/blob/f2d1adc96a7e25e2e61425f5fa97a8f7a7225e77/nanovllm/qwen38/server.py", + "commit": "f2d1adc96a7e25e2e61425f5fa97a8f7a7225e77", + "paths": ["nanovllm/qwen38/server.py"], + "content_sha256": "6a9c8c60f688705a722c1f269ea5aad7ede9d4a38f0251e580f24773846780cb", + "issue": "https://github.com/InferMatrixLab/InferRSI/issues/67", + "note": "handle_error and log_message are no-ops. The file comment is the sentence quoted in the case. This blob does not contain a captured HTTP 400 body." + }, + { + "type": "git", + "url": "https://github.com/jiangkuaixue123/nano-vllm/blob/3f7516a7b4a2c0a34871a453c294bc23b1b6c67c/nanovllm/qwen38/runner.py", + "commit": "3f7516a7b4a2c0a34871a453c294bc23b1b6c67c", + "paths": ["nanovllm/qwen38/runner.py"], + "content_sha256": "3b99c999c887f6acc2f2b6aa957c01680246fe424fb3da961322c06dacf857a0", + "issue": "https://github.com/InferMatrixLab/InferRSI/issues/67", + "note": "graph_headroom = 256 and bucket_for(prompt length + headroom)." + }, + { + "type": "document", + "url": "https://github.com/jiangkuaixue123/nano-vllm/pull/8", + "version": "2026-10-22T17:05:30Z", + "sha256": "b4c5b6c2c3681976df2ccc7ea41c9b5c245aa0769d2c82e91e26f4eca8cc1684", + "issue": "https://github.com/InferMatrixLab/InferRSI/issues/67", + "run": "1951", + "note": "Merged PR body: the official client reports success=true for the streamed error; job 1951 errored after 876 tokens at the graph capacity boundary. Digest is that body retrieved 2026-10-31." + }, + { + "type": "git", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5e17195280249cd0bbd5ef87d79d8654cce7056/evidence/qwen38-h20-fp8-single-request-vllm032/results/full-context-graph/preparation.md", + "commit": "d5e17195280249cd0bbd5ef87d79d8654cce7056", + "paths": ["evidence/qwen38-h20-fp8-single-request-vllm032/results/full-context-graph/preparation.md"], + "content_sha256": "1f186a3eb8e2ee6c20026848799c1801d7a15c91088724d2f86bcc18a61d95a6", + "issue": "https://github.com/InferMatrixLab/InferRSI/issues/67", + "note": "Retained on main: prompt+256 headroom demotes an 8192-input/2048-output request after its reserved decode window. Campaign acceptance remains INCONCLUSIVE." + } + ], + "applicability": [ + "OpenAI-compatible HTTP wrapper over a custom inference engine", + "Bucketed CUDA Graph decode with a headroom-selected bucket" + ], + "tags": ["serving", "error-handling", "observability", "capacity", "silent-failure"], + "limitations": [ + "One engine, one card", + "Behaviour past capacity not characterised beyond the failure itself", + "No concurrency or multi-client qualification", + "The located public artifacts do not contain the literal failure index 993; the retained capacity-boundary run is job 1951 at 876 tokens", + "The empty-log mechanism is the cited server.py; a captured HTTP 400 empty body was not in that blob", + "preparation.md is a later retained note of the headroom rule, not a new measurement for this case" + ], + "attribution": "nano-vllm server.py at f2d1adc96a7e25e2e61425f5fa97a8f7a7225e77, runner.py at 3f7516a7b4a2c0a34871a453c294bc23b1b6c67c, nano-vllm PR 8, and InferRSI evidence preparation.md at d5e17195280249cd0bbd5ef87d79d8654cce7056" } diff --git a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md index bb5b17707..fcaa1ca54 100644 --- a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md +++ b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md @@ -26,3 +26,14 @@ device. The third row is specific to bucketed graph execution. Limitations: observed on one engine and one card; the exact fallback behaviour past capacity was not characterised beyond "generation failed". + +Retained records: the empty-log handler is `nanovllm/qwen38/server.py` at +`f2d1adc96a7e25e2e61425f5fa97a8f7a7225e77`. `graph_headroom = 256` is +`nanovllm/qwen38/runner.py` at `3f7516a7b4a2c0a34871a453c294bc23b1b6c67c`. +[nano-vllm pull request 8](https://github.com/jiangkuaixue123/nano-vllm/pull/8) +(body sha256 `b4c5b6c2c3681976df2ccc7ea41c9b5c245aa0769d2c82e91e26f4eca8cc1684`) +records a streamed error reported as success and job 1951 failing after 876 +tokens. Prompt+256 headroom on an 8192-input/2048-output request is retained in +`evidence/qwen38-h20-fp8-single-request-vllm032/results/full-context-graph/preparation.md` +at `d5e17195280249cd0bbd5ef87d79d8654cce7056`. The literal index 993 is not in +those artifacts. diff --git a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.json b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.json index bf2a8a28b..c073c525d 100644 --- a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.json +++ b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.json @@ -1,26 +1,52 @@ { - "id": "case.qwen38-upstream-fp8-kernels-vs-own-triton", - "kind": "case", - "origin": ["first_party"], - "state": "historical_observation", - "content": "qwen38-upstream-fp8-kernels-vs-own-triton.md", - "sources": [ - { - "type": "local", - "note": "Qwen3.8 dedicated-engine campaign, FP8 kernel microbenchmarks", - "paths": ["evidence/"] - } - ], - "applicability": [ - "SM90-class accelerator, FP8 e4m3 with 128x128 block scales", - "M in the thousands (prefill), batch size 1 decode measured separately", - "torch 2.14 engine environment" - ], - "tags": ["fp8", "gemm", "build-vs-borrow", "microbenchmark", "licensing"], - "limitations": [ - "Single card; numbers are microbenchmarks, not end-to-end", - "Thresholds registered for one shape family and one device", - "Upstream versions may change both performance and licensing" - ], - "attribution": "Dedicated-engine campaign; measurements taken on the target card" + "id": "case.qwen38-upstream-fp8-kernels-vs-own-triton", + "kind": "case", + "origin": ["first_party"], + "state": "historical_observation", + "content": "qwen38-upstream-fp8-kernels-vs-own-triton.md", + "sources": [ + { + "type": "internal", + "issue": "https://github.com/InferMatrixLab/InferRSI/issues/67", + "run": "issuecomment-6336720892", + "artifact": "https://github.com/InferMatrixLab/InferRSI/issues/67#issuecomment-6336720892", + "version": "2026-10-22T08:50:35Z", + "sha256": "37b84900b48fc5da201792c280519969514d2bbed0afe75a3c4ef56329077201", + "integrity_status": "issue_comment_body_sha256", + "note": "Ledger v7: own Triton 2.8 TFLOPS, FlashInfer 273 TFLOPS and 3273 GB/s at M=1, DeepGEMM 279 TFLOPS, go line 280, 2.5% shortfall, 18 KB deep_gemm placeholder, vendored extension pinned to a torch minor. Digest is the comment body retrieved 2026-10-31." + }, + { + "type": "document", + "url": "https://github.com/jiangkuaixue123/nano-vllm/pull/3", + "version": "2026-10-23T07:20:41Z", + "sha256": "172d437e2772a4e5aefff1f9a03a4fd381f6b11e0c12e0821eb2541c5c568c34", + "issue": "https://github.com/InferMatrixLab/InferRSI/issues/67", + "note": "Unmerged PR body: native block-FP8 kernel 2.8 TFLOPS at M=8192 and 8K prefill 145 s. Digest is the body retrieved 2026-10-31. The ledger later marks 145.3 s as including first-call JIT." + }, + { + "type": "internal", + "issue": "https://github.com/InferMatrixLab/InferRSI/issues/67", + "run": "issuecomment-6336567189", + "artifact": "https://github.com/InferMatrixLab/InferRSI/issues/67#issuecomment-6336567189", + "version": "2026-10-22T08:42:07Z", + "sha256": "aedccaca1c8fbdb8c09eac5a79f9a48f43ec960ce33f4513822648c10b98a3dd", + "integrity_status": "issue_comment_body_sha256", + "note": "Pre-registered go line 280 TFLOPS against a 279 measurement (0.36%). The following ledger comment supersedes the adoption decision with FlashInfer at 273." + } + ], + "applicability": [ + "SM90-class accelerator, FP8 e4m3 with 128x128 block scales", + "M in the thousands (prefill), batch size 1 decode measured separately", + "torch 2.14 engine environment" + ], + "tags": ["fp8", "gemm", "build-vs-borrow", "microbenchmark", "licensing"], + "limitations": [ + "Single card; numbers are microbenchmarks, not end-to-end", + "Thresholds registered for one shape family and one device", + "Upstream versions may change both performance and licensing", + "The located ledger records go >=280 TFLOPS and does not contain an abandon-below-250 threshold", + "145 s is in the unmerged PR body; the same day's ledger withdraws 145.3 s as evidence because it includes first-call JIT", + "No raw microbenchmark file for these TFLOPS figures was in the retained evidence tree; the digests above are the public ledger bodies" + ], + "attribution": "Issue #67 comments 6336720892 and 6336567189, and nano-vllm pull request 3; microbenchmarks, not an end-to-end claim" } diff --git a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.md b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.md index 802d2f1c6..ab2b81f77 100644 --- a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.md +++ b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.md @@ -34,3 +34,14 @@ engine environment. The M=1 numbers are microbenchmarks, not end-to-end decode. Limitations: single-card microbenchmarks; end-to-end prefill was measured separately; upstream versions may change both the numbers and the licensing. + +Retained records: the 2.8 / 273 / 279 TFLOPS table, 3273 GB/s at M=1, the 280 +TFLOPS go line, the 2.5% shortfall, the placeholder package and the torch pin +are Issue #67 comment 6336720892 (body sha256 +`37b84900b48fc5da201792c280519969514d2bbed0afe75a3c4ef56329077201`). 2.8 TFLOPS +at M=8192 and the 145 s prefill are also in +[nano-vllm pull request 3](https://github.com/jiangkuaixue123/nano-vllm/pull/3) +(body sha256 `172d437e2772a4e5aefff1f9a03a4fd381f6b11e0c12e0821eb2541c5c568c34`); +the same day's ledger withdraws 145.3 s because it includes first-call JIT. +Comment 6336567189 records 279 against the 280 line. The located ledger does +not contain an abandon-below-250 threshold. From a5b99d35fbffa2cf0ee0479082ab6d8d68e79d2e Mon Sep 17 00:00:00 2001 From: jiangkuaixue123 Date: Sun, 1 Nov 2026 01:09:25 +0000 Subject: [PATCH 3/8] docs(casebook): align SSE and FP8 case wording with retained records Split stream checks so a normal stop finish is success, and reserve an exact length assertion for a capacity fixture. Label the 145 s prefill as withdrawn JIT-confounded timing and drop the unsupported 250 threshold. Agent-assisted-by: Cursor / grok-4.7 [effort=unknown] (author) Signed-off-by: jiangkuaixue123 Co-authored-by: jiangkuaixue123 --- ...en38-silent-capacity-and-error-failures.md | 5 +-- ...38-upstream-fp8-kernels-vs-own-triton.json | 5 +-- ...en38-upstream-fp8-kernels-vs-own-triton.md | 31 ++++++++++--------- 3 files changed, 23 insertions(+), 18 deletions(-) diff --git a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md index fcaa1ca54..c760e7246 100644 --- a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md +++ b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md @@ -7,7 +7,7 @@ each cost a debugging round. | Symptom | What actually happened | Detection that would have caught it | | --- | --- | --- | -| HTTP 200 with an SSE `error` frame mid-stream | Decode ran past the configured graph bucket, the engine fell back or failed, and the wrapper kept the stream open | Assert `completion_tokens` equals the requested count and `finish_reason == length` on every request | +| HTTP 200 with an SSE `error` frame mid-stream | Decode ran past the configured graph bucket, the engine fell back or failed, and the wrapper kept the stream open | On every request, parse each SSE frame, reject an embedded `error` or a truncated stream, and require a terminal frame whose `finish_reason` is `stop` or `length`. When usage is present, check `completion_tokens` against it. A normal EOS/`stop` finish below `max_tokens` is success. Assert `completion_tokens` equals a frozen budget and `finish_reason == length` only on a capacity fixture that disables early stop | | HTTP 400, empty body, 0-byte server log | The error handler deliberately dropped all exceptions ("client addresses and arbitrary exceptions do not belong in service logs") | Log a structured line with exception type and a truncated message; never return an error status without a body | | Generation failed at token 993 with no message | The graph bucket was chosen from `prompt + headroom` with `headroom = 256`, while the request asked for 2048 output tokens | Validate bucket coverage against maximum output length, not against prompt slack, before starting the run | @@ -15,7 +15,8 @@ Disposition: a serving layer must never convert an internal failure into a success response. Log the exception type to stderr even when the payload must be redacted, return a JSON error body for every 4xx/5xx, and treat capacity (bucket, context, output length) as a pre-flight assertion rather than a runtime -surprise. +surprise. `max_tokens` is an upper bound. Do not treat a shorter `stop` finish +as a failed length request. Revisit condition: any wrapper that owns stream framing or error translation. First check: send one request whose output length exceeds the configured diff --git a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.json b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.json index c073c525d..cacf24be8 100644 --- a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.json +++ b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.json @@ -44,8 +44,9 @@ "Single card; numbers are microbenchmarks, not end-to-end", "Thresholds registered for one shape family and one device", "Upstream versions may change both performance and licensing", - "The located ledger records go >=280 TFLOPS and does not contain an abandon-below-250 threshold", - "145 s is in the unmerged PR body; the same day's ledger withdraws 145.3 s as evidence because it includes first-call JIT", + "The case body and this sidecar both keep the recorded go line of >=280 TFLOPS and the observed 273/279 TFLOPS figures", + "An abandon-below-250 threshold is absent from the located records and is not asserted", + "145 s remains only as a withdrawn, JIT-confounded report from the unmerged PR body; it is not timing evidence", "No raw microbenchmark file for these TFLOPS figures was in the retained evidence tree; the digests above are the public ledger bodies" ], "attribution": "Issue #67 comments 6336720892 and 6336567189, and nano-vllm pull request 3; microbenchmarks, not an end-to-end claim" diff --git a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.md b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.md index ab2b81f77..3a42ed9d4 100644 --- a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.md +++ b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.md @@ -2,8 +2,10 @@ A dedicated engine shipped a Triton block-scaled FP8 GEMM that padded 32 singleton lanes into an M64 tile to keep numerics stable. It ran at 2.8 TFLOPS -on H20-3e and made an 8K prefill take 145 s. Measuring the upstream libraries on -the same shapes turned the build-versus-borrow decision into arithmetic. +on H20-3e. An 8K prefill was also reported at 145 s; the same day's ledger +withdraws 145.3 s because that sample includes first-call JIT and is not timing +evidence. Measuring the upstream libraries on the same shapes turned the +build-versus-borrow decision into arithmetic. | Implementation | 8192x17408x5120 | Note | | --- | --- | --- | @@ -11,10 +13,12 @@ the same shapes turned the build-versus-borrow decision into arithmetic. | FlashInfer block-scaled (installed from PyPI, Apache-2.0) | 273 TFLOPS | usable in the engine environment | | DeepGEMM (vendored inside vLLM, prebuilt `.so`) | 279 TFLOPS | unusable outside vLLM's environment: the extension pins a torch minor version | -Thresholds were registered before the measurement: go at >=280 TFLOPS, abandon -below 250. FlashInfer measured 273, so it was adopted with the 2.5% shortfall -recorded rather than rounded away. The same library reports 3273 GB/s for M=1, -which makes it a candidate for the bandwidth-bound decode path as well. +The located ledger registers one go line before the measurement: >=280 TFLOPS. +FlashInfer measured 273 and DeepGEMM measured 279, so FlashInfer was adopted +with the 2.5% shortfall recorded rather than rounded away. An abandon-below-250 +line is not in those records, so this case does not treat 250 as a registered +threshold. The same library reports 3273 GB/s for M=1, which makes it a +candidate for the bandwidth-bound decode path as well. Two traps worth recording. `pip install deep_gemm` resolves to an unrelated 18 KB placeholder on PyPI; the real project is source-only. And a vendored @@ -22,8 +26,9 @@ extension compiled against one torch minor version fails at import on another, so "reuse the dependency of the reference server" is not a shortcut. Disposition: measure the upstream implementation on your own shapes before -writing a kernel whose only claim is numerical caution. Register the go/abandon -threshold first; record the shortfall if you adopt below it. +writing a kernel whose only claim is numerical caution. Register the go line +first; record the shortfall if you adopt below it. Do not cite the withdrawn +145 s sample as prefill timing. Revisit condition: new shapes, a different accelerator, or a new upstream release. First check: rerun the shape sweep and confirm the largest shapes still @@ -38,10 +43,8 @@ separately; upstream versions may change both the numbers and the licensing. Retained records: the 2.8 / 273 / 279 TFLOPS table, 3273 GB/s at M=1, the 280 TFLOPS go line, the 2.5% shortfall, the placeholder package and the torch pin are Issue #67 comment 6336720892 (body sha256 -`37b84900b48fc5da201792c280519969514d2bbed0afe75a3c4ef56329077201`). 2.8 TFLOPS -at M=8192 and the 145 s prefill are also in +`37b84900b48fc5da201792c280519969514d2bbed0afe75a3c4ef56329077201`). The +withdrawn 145 s report, not a timing result, is also in [nano-vllm pull request 3](https://github.com/jiangkuaixue123/nano-vllm/pull/3) -(body sha256 `172d437e2772a4e5aefff1f9a03a4fd381f6b11e0c12e0821eb2541c5c568c34`); -the same day's ledger withdraws 145.3 s because it includes first-call JIT. -Comment 6336567189 records 279 against the 280 line. The located ledger does -not contain an abandon-below-250 threshold. +(body sha256 `172d437e2772a4e5aefff1f9a03a4fd381f6b11e0c12e0821eb2541c5c568c34`). +Comment 6336567189 records 279 against the 280 line and does not register 250. From 591f3b613e688a1ee3bbdae518811ad7300a44fd Mon Sep 17 00:00:00 2001 From: jiangkuaixue123 Date: Sun, 1 Nov 2026 01:11:16 +0000 Subject: [PATCH 4/8] docs(casebook): regenerate the skill discovery projection Refresh the packaged index from the rebased casebook source so the installed digest matches the current skill bytes. Agent-assisted-by: Cursor / grok-4.7 [effort=unknown] (author) Signed-off-by: jiangkuaixue123 Co-authored-by: jiangkuaixue123 --- .../skills/inference-optimization-casebook.json | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-) diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/inference-optimization-casebook.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/inference-optimization-casebook.json index 344bae3cb..60aee5636 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/inference-optimization-casebook.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/inference-optimization-casebook.json @@ -33,10 +33,16 @@ "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen-ascend-rejected-directions.md", "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen-b300-rejected-directions.json", "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen-b300-rejected-directions.md", + "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.json", + "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.md", + "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.json", + "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md", + "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.json", + "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.md", "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/metadata.json", "src/inferrsi/data/skill_library/skill-index.json" ], - "content_sha256": "89a6a82c692f7f5192f4d7f4c3a5ff08a4df3b1c4bdaa0aafb7e04292d8aa5b4" + "content_sha256": "e6ff2de3835ad455df1c5bab7ad002d8489d8c7ef9abeab9660206f602cc8fe7" } ], "index_type": "packaged-skill", @@ -45,7 +51,7 @@ }, "checkout_ref": "checkout:inference-optimization-casebook", "source_files": { - "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/SKILL.md": "89a6a82c692f7f5192f4d7f4c3a5ff08a4df3b1c4bdaa0aafb7e04292d8aa5b4", + "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/SKILL.md": "e6ff2de3835ad455df1c5bab7ad002d8489d8c7ef9abeab9660206f602cc8fe7", "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/admission-drain-occupancy.json": "3319e23bbe2bd40e2d45fb89dc87bdc92d4df2b5ddd8086fcf84813b43885b7c", "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/admission-drain-occupancy.md": "e88bbac5d253a8e7e3a3d5e21047f23f2970c11ac23887d4303ab94fca52dd27", "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/drama3-rejected-directions.json": "2cc0a526fe0509551607a03e7626dd3d7844e643c4afb194837f75cddead8b12", @@ -54,6 +60,12 @@ "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen-ascend-rejected-directions.md": "be396983812262515899a3e2ba4f01b23ff92e1b836881882c7580372b8ae552", "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen-b300-rejected-directions.json": "146ea1564a46ed03092c7dc5796b318fc05b04b1821105ee00eb7fa7fa1333e2", "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen-b300-rejected-directions.md": "cce2930f544a0f78f21797fba4d8ded298e4a81f76c34d7977b0e31d361bc738", + "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.json": "13cae252f2abf8fe364132f285a84abbb84c785283322ef5c55dec1f9d8bc181", + "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.md": "73e89ce9ac408710f8f810c28d09b1b5697c09b4837923b8cc4ddd1cdb5ba92f", + "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.json": "c72d5555bc6dc9d6d63b4aaa767f2f52b4c3d2a6de7008f578de4b68c9a2e692", + "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md": "8197140a701efc52b1c7f0b6df1cb4d840843b3eb4d1457a6d36a9a396eb90b2", + "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.json": "97db326a4671cec41bc2141381c3e093d761f140481022abe9116b0d9c163628", + "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.md": "ea61515741897ebcfdcc0878dd5e98d32362682d0ce5294bbb45c98600815cc0", "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/metadata.json": "963488658f327d2379183d70357eabf832d70a6a70db2aa67f548493f4088d97", "src/inferrsi/data/skill_library/skill-index.json": "9ea8b2ef190777e683a3905f0614612a37782be4d655d230c000188b0c7073db" }, From e1a2292a935cfc46a0451211ee059b22cd9b2db7 Mon Sep 17 00:00:00 2001 From: jiangkuaixue123 Date: Sun, 1 Nov 2026 01:17:57 +0000 Subject: [PATCH 5/8] docs(casebook): cite case records without copying source digests Link each Qwen3.8 case to its typed record, keep digests in that record, and refresh the coverage view for the new case bytes. Agent-assisted-by: Cursor / grok-4.7 [effort=unknown] (author) Signed-off-by: jiangkuaixue123 Co-authored-by: jiangkuaixue123 --- ...n38-cuda-graph-state-not-accumulating.json | 1 + ...wen38-cuda-graph-state-not-accumulating.md | 19 ++++++++------- ...38-silent-capacity-and-error-failures.json | 1 + ...en38-silent-capacity-and-error-failures.md | 24 +++++++++++-------- ...38-upstream-fp8-kernels-vs-own-triton.json | 1 + ...en38-upstream-fp8-kernels-vs-own-triton.md | 16 +++++++------ tests/test_knowledge_contract.py | 2 +- 7 files changed, 38 insertions(+), 26 deletions(-) diff --git a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.json b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.json index df7fdd05a..bbc3123b0 100644 --- a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.json +++ b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.json @@ -4,6 +4,7 @@ "origin": ["first_party"], "state": "historical_observation", "content": "qwen38-cuda-graph-state-not-accumulating.md", + "required_resources": ["qwen38-cuda-graph-state-not-accumulating.json"], "sources": [ { "type": "internal", diff --git a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.md b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.md index 5206f20c8..f9786fde3 100644 --- a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.md +++ b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.md @@ -1,5 +1,8 @@ # CUDA Graph replay read capture-time state: ids stayed equal while logits diverged +Kind, origin, state and source identities belong to the +[typed source and scope record](qwen38-cuda-graph-state-not-accumulating.json). + A single-request Qwen3.8 decode engine (64 layers: 48 linear-attention/DeltaNet plus 16 full-attention) captured its whole decode step into a CUDA Graph. Greedy token ids matched the eager path exactly, so three successive regression gates @@ -33,11 +36,11 @@ no cross-device qualification. The configuration was recorded as INCONCLUSIVE for acceptance purposes because the measurement that exposed it happened after the timing run. -Retained records: the 0.70 gap is Issue #67 comment 6337245823 (body sha256 -`de68a42eb35593b850c63e9bdc29f6896775d8f15fa34fdc8f8be8a6866a7f3f`). Comment -6331682106 names `store:case.qwen38-graph-state-not-accumulating-cand` and a -tiny-fixture bitwise pass (body sha256 -`9f7d7085e4d4508712fa5f5a1022111092ad897703663c697773e20bcccd0663`); it does not -contain the literal `0.00e+00`, and prefix `5c7700d` is not a public commit. -The in-place buffers and dedicated capture pool are `nanovllm/qwen38/graph.py` -at `3f7516a7b4a2c0a34871a453c294bc23b1b6c67c`. +Retained records: [Issue #67](https://github.com/InferMatrixLab/InferRSI/issues/67) +comment 6337245823 records the 0.70 logit gap with identical greedy ids. +Comment 6331682106 names `store:case.qwen38-graph-state-not-accumulating-cand` +and a tiny-fixture bitwise pass; it does not contain the literal `0.00e+00`, +and prefix `5c7700d` is not a public commit. The in-place buffers and dedicated +capture pool are +[nanovllm/qwen38/graph.py](https://github.com/jiangkuaixue123/nano-vllm/blob/3f7516a7b4a2c0a34871a453c294bc23b1b6c67c/nanovllm/qwen38/graph.py). +Digests stay in the typed record. diff --git a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.json b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.json index ab29841e0..4868f9b31 100644 --- a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.json +++ b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.json @@ -4,6 +4,7 @@ "origin": ["first_party"], "state": "historical_observation", "content": "qwen38-silent-capacity-and-error-failures.md", + "required_resources": ["qwen38-silent-capacity-and-error-failures.json"], "sources": [ { "type": "git", diff --git a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md index c760e7246..e93b7c9d4 100644 --- a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md +++ b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md @@ -1,5 +1,8 @@ # Three ways an inference server reported success while failing +Kind, origin, state and source identities belong to the +[typed source and scope record](qwen38-silent-capacity-and-error-failures.json). + While measuring a single-request engine through its OpenAI-compatible HTTP endpoint, three distinct failures produced no actionable evidence: two returned a success status, one returned an empty error body. Each is cheap to detect and @@ -28,13 +31,14 @@ device. The third row is specific to bucketed graph execution. Limitations: observed on one engine and one card; the exact fallback behaviour past capacity was not characterised beyond "generation failed". -Retained records: the empty-log handler is `nanovllm/qwen38/server.py` at -`f2d1adc96a7e25e2e61425f5fa97a8f7a7225e77`. `graph_headroom = 256` is -`nanovllm/qwen38/runner.py` at `3f7516a7b4a2c0a34871a453c294bc23b1b6c67c`. -[nano-vllm pull request 8](https://github.com/jiangkuaixue123/nano-vllm/pull/8) -(body sha256 `b4c5b6c2c3681976df2ccc7ea41c9b5c245aa0769d2c82e91e26f4eca8cc1684`) -records a streamed error reported as success and job 1951 failing after 876 -tokens. Prompt+256 headroom on an 8192-input/2048-output request is retained in -`evidence/qwen38-h20-fp8-single-request-vllm032/results/full-context-graph/preparation.md` -at `d5e17195280249cd0bbd5ef87d79d8654cce7056`. The literal index 993 is not in -those artifacts. +Retained records, with digests in the typed record: +[Issue #67](https://github.com/InferMatrixLab/InferRSI/issues/67); +the empty-log handler in +[nanovllm/qwen38/server.py](https://github.com/jiangkuaixue123/nano-vllm/blob/f2d1adc96a7e25e2e61425f5fa97a8f7a7225e77/nanovllm/qwen38/server.py); +`graph_headroom = 256` in +[nanovllm/qwen38/runner.py](https://github.com/jiangkuaixue123/nano-vllm/blob/3f7516a7b4a2c0a34871a453c294bc23b1b6c67c/nanovllm/qwen38/runner.py); +[nano-vllm pull request 8](https://github.com/jiangkuaixue123/nano-vllm/pull/8), +where job 1951 fails after 876 tokens and a streamed error is reported as +success; and the later headroom note in +[preparation.md](https://github.com/InferMatrixLab/InferRSI/blob/d5e17195280249cd0bbd5ef87d79d8654cce7056/evidence/qwen38-h20-fp8-single-request-vllm032/results/full-context-graph/preparation.md). +The literal index 993 is not in those artifacts. diff --git a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.json b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.json index cacf24be8..b01de32a5 100644 --- a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.json +++ b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.json @@ -4,6 +4,7 @@ "origin": ["first_party"], "state": "historical_observation", "content": "qwen38-upstream-fp8-kernels-vs-own-triton.md", + "required_resources": ["qwen38-upstream-fp8-kernels-vs-own-triton.json"], "sources": [ { "type": "internal", diff --git a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.md b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.md index 3a42ed9d4..a30af27ea 100644 --- a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.md +++ b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.md @@ -1,5 +1,8 @@ # A self-written block-scaled FP8 GEMM was 100x slower than the upstream kernel +Kind, origin, state and source identities belong to the +[typed source and scope record](qwen38-upstream-fp8-kernels-vs-own-triton.json). + A dedicated engine shipped a Triton block-scaled FP8 GEMM that padded 32 singleton lanes into an M64 tile to keep numerics stable. It ran at 2.8 TFLOPS on H20-3e. An 8K prefill was also reported at 145 s; the same day's ledger @@ -40,11 +43,10 @@ engine environment. The M=1 numbers are microbenchmarks, not end-to-end decode. Limitations: single-card microbenchmarks; end-to-end prefill was measured separately; upstream versions may change both the numbers and the licensing. -Retained records: the 2.8 / 273 / 279 TFLOPS table, 3273 GB/s at M=1, the 280 -TFLOPS go line, the 2.5% shortfall, the placeholder package and the torch pin -are Issue #67 comment 6336720892 (body sha256 -`37b84900b48fc5da201792c280519969514d2bbed0afe75a3c4ef56329077201`). The -withdrawn 145 s report, not a timing result, is also in -[nano-vllm pull request 3](https://github.com/jiangkuaixue123/nano-vllm/pull/3) -(body sha256 `172d437e2772a4e5aefff1f9a03a4fd381f6b11e0c12e0821eb2541c5c568c34`). +Retained records, with digests in the typed record: the 2.8 / 273 / 279 TFLOPS +table, 3273 GB/s at M=1, the 280 TFLOPS go line, the 2.5% shortfall, the +placeholder package and the torch pin are +[Issue #67](https://github.com/InferMatrixLab/InferRSI/issues/67) comment +6336720892. The withdrawn 145 s report is in +[nano-vllm pull request 3](https://github.com/jiangkuaixue123/nano-vllm/pull/3). Comment 6336567189 records 279 against the 280 line and does not register 250. diff --git a/tests/test_knowledge_contract.py b/tests/test_knowledge_contract.py index f0211a8f3..f1dbb9157 100644 --- a/tests/test_knowledge_contract.py +++ b/tests/test_knowledge_contract.py @@ -58,7 +58,7 @@ def _case_records(): def test_case_documents_cite_their_record_without_copying_digests(): cases = list(_case_records()) - assert len(cases) == 7 + assert len(cases) == 10 for path, record in cases: body = (path.parent / record["content"]).read_text(encoding="utf-8") # The JSON record is the single home for source identities and digests; From 0d1bbccb2ed5f4622ad848b077d30dbddc19eacb Mon Sep 17 00:00:00 2001 From: jiangkuaixue123 Date: Sun, 1 Nov 2026 01:18:03 +0000 Subject: [PATCH 6/8] docs(casebook): regenerate the skill discovery projection Point the packaged index at the rebased casebook source after the SSE and FP8 wording alignment. Agent-assisted-by: Cursor / grok-4.7 [effort=unknown] (author) Signed-off-by: jiangkuaixue123 Co-authored-by: jiangkuaixue123 --- .../skills/inference-optimization-casebook.json | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/inference-optimization-casebook.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/inference-optimization-casebook.json index 60aee5636..384974e45 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/inference-optimization-casebook.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/inference-optimization-casebook.json @@ -60,12 +60,12 @@ "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen-ascend-rejected-directions.md": "be396983812262515899a3e2ba4f01b23ff92e1b836881882c7580372b8ae552", "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen-b300-rejected-directions.json": "146ea1564a46ed03092c7dc5796b318fc05b04b1821105ee00eb7fa7fa1333e2", "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen-b300-rejected-directions.md": "cce2930f544a0f78f21797fba4d8ded298e4a81f76c34d7977b0e31d361bc738", - "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.json": "13cae252f2abf8fe364132f285a84abbb84c785283322ef5c55dec1f9d8bc181", - "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.md": "73e89ce9ac408710f8f810c28d09b1b5697c09b4837923b8cc4ddd1cdb5ba92f", - "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.json": "c72d5555bc6dc9d6d63b4aaa767f2f52b4c3d2a6de7008f578de4b68c9a2e692", - "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md": "8197140a701efc52b1c7f0b6df1cb4d840843b3eb4d1457a6d36a9a396eb90b2", - "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.json": "97db326a4671cec41bc2141381c3e093d761f140481022abe9116b0d9c163628", - "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.md": "ea61515741897ebcfdcc0878dd5e98d32362682d0ce5294bbb45c98600815cc0", + "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.json": "7d02a7435560745d642dc1137859e2a56914327f1235c52e726b277fdf470a27", + "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.md": "8a51ed2c55d0aebff87fddfba8eb9cfbb415e014f7534d3fe58cbd72e8f5deb5", + "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.json": "54b686f7a1f935ace148c49ee7ec343343dc38877fe9d5a6efadf46ef1a6e27e", + "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md": "2f53bb0559231cc43da592fc47065209143e9d85e5b67eb129126f53c777953f", + "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.json": "07fbd0ff59a1fff168f48b7e8084f038bd855f9cb493913acc850bf7d492abd0", + "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.md": "7b01b9719265430ed2be83e0535def09af5393035c714041fb16e6387ed0d6df", "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/metadata.json": "963488658f327d2379183d70357eabf832d70a6a70db2aa67f548493f4088d97", "src/inferrsi/data/skill_library/skill-index.json": "9ea8b2ef190777e683a3905f0614612a37782be4d655d230c000188b0c7073db" }, From 405033073d1140084da654c3701ac32cc4d1e4a1 Mon Sep 17 00:00:00 2001 From: jiangkuaixue123 Date: Mon, 2 Nov 2026 07:35:16 +0000 Subject: [PATCH 7/8] docs(casebook): cite job 1951 at 876 tokens for the capacity failure The retained run is job 1951, which failed after 876 tokens. Drop the unverified token-993 figure from the skill summary and the case. Agent-assisted-by: Cursor / grok-4.7 [effort=unknown] (author) Signed-off-by: jiangkuaixue123 Co-authored-by: jiangkuaixue123 --- .../methods/inference-optimization-casebook/SKILL.md | 2 +- .../cases/qwen38-silent-capacity-and-error-failures.json | 2 +- .../cases/qwen38-silent-capacity-and-error-failures.md | 3 +-- 3 files changed, 3 insertions(+), 4 deletions(-) diff --git a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/SKILL.md b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/SKILL.md index 851c977f9..bc40b9078 100644 --- a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/SKILL.md +++ b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/SKILL.md @@ -89,7 +89,7 @@ Current owner-local records retain rejected directions alongside their limits: - [CUDA Graph state not accumulating](cases/qwen38-cuda-graph-state-not-accumulating.md): token ids matched while per-step logits diverged, so ids are not a correctness oracle. - [Silent capacity and error failures](cases/qwen38-silent-capacity-and-error-failures.md): - HTTP 200 with an error frame, HTTP 400 with an empty body, failure at token 993. + HTTP 200 with an error frame, HTTP 400 with an empty body, job 1951 failed after 876 tokens. - [Upstream FP8 kernels vs own Triton](cases/qwen38-upstream-fp8-kernels-vs-own-triton.md): a numerically cautious kernel ran 100x slower than the borrowed library. diff --git a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.json b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.json index 4868f9b31..507c7045f 100644 --- a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.json +++ b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.json @@ -52,7 +52,7 @@ "One engine, one card", "Behaviour past capacity not characterised beyond the failure itself", "No concurrency or multi-client qualification", - "The located public artifacts do not contain the literal failure index 993; the retained capacity-boundary run is job 1951 at 876 tokens", + "The retained capacity-boundary observation is job 1951, which failed after 876 tokens", "The empty-log mechanism is the cited server.py; a captured HTTP 400 empty body was not in that blob", "preparation.md is a later retained note of the headroom rule, not a new measurement for this case" ], diff --git a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md index e93b7c9d4..e65e85bd9 100644 --- a/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md +++ b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md @@ -12,7 +12,7 @@ each cost a debugging round. | --- | --- | --- | | HTTP 200 with an SSE `error` frame mid-stream | Decode ran past the configured graph bucket, the engine fell back or failed, and the wrapper kept the stream open | On every request, parse each SSE frame, reject an embedded `error` or a truncated stream, and require a terminal frame whose `finish_reason` is `stop` or `length`. When usage is present, check `completion_tokens` against it. A normal EOS/`stop` finish below `max_tokens` is success. Assert `completion_tokens` equals a frozen budget and `finish_reason == length` only on a capacity fixture that disables early stop | | HTTP 400, empty body, 0-byte server log | The error handler deliberately dropped all exceptions ("client addresses and arbitrary exceptions do not belong in service logs") | Log a structured line with exception type and a truncated message; never return an error status without a body | -| Generation failed at token 993 with no message | The graph bucket was chosen from `prompt + headroom` with `headroom = 256`, while the request asked for 2048 output tokens | Validate bucket coverage against maximum output length, not against prompt slack, before starting the run | +| Job 1951 failed after 876 tokens with no message | The graph bucket was chosen from `prompt + headroom` with `headroom = 256`, while the request asked for 2048 output tokens | Validate bucket coverage against maximum output length, not against prompt slack, before starting the run | Disposition: a serving layer must never convert an internal failure into a success response. Log the exception type to stderr even when the payload must be @@ -41,4 +41,3 @@ the empty-log handler in where job 1951 fails after 876 tokens and a streamed error is reported as success; and the later headroom note in [preparation.md](https://github.com/InferMatrixLab/InferRSI/blob/d5e17195280249cd0bbd5ef87d79d8654cce7056/evidence/qwen38-h20-fp8-single-request-vllm032/results/full-context-graph/preparation.md). -The literal index 993 is not in those artifacts. From dc5a0a3e2cd0303d00391736641a6fd7f18bc602 Mon Sep 17 00:00:00 2001 From: jiangkuaixue123 Date: Mon, 2 Nov 2026 07:35:28 +0000 Subject: [PATCH 8/8] docs(casebook): regenerate the skill discovery projection Refresh the packaged index and coverage view from the job 1951 wording. Agent-assisted-by: Cursor / grok-4.7 [effort=unknown] (author) Signed-off-by: jiangkuaixue123 Co-authored-by: jiangkuaixue123 --- docs/skill-coverage.md | 2 +- .../data/repo_packs/inferrsi/knowledge/README.md | 4 ++-- .../skills/ascend-aclgraph-capture-replay.json | 4 ++-- .../ascend/skills/ascend-device-health.json | 4 ++-- .../ascend/skills/ascend-distributed-parallel.json | 4 ++-- .../knowledge/ascend/skills/ascend-graph-mode.json | 4 ++-- .../ascend/skills/ascend-model-adaptation.json | 4 ++-- .../ascend/skills/ascend-operator-profiling.json | 4 ++-- .../ascend/skills/ascend-operator-tuning.json | 4 ++-- .../ascend/skills/ascend-optimization-casebook.json | 4 ++-- .../ascend/skills/ascend-performance-analysis.json | 4 ++-- .../skills/ascend-video-diffusion-optimization.json | 4 ++-- .../ascend/skills/ascendc-tiling-design.json | 4 ++-- .../ascend/skills/cann-attribution-overlap.json | 4 ++-- .../inferrsi/knowledge/ascend/skills/torch-npu.json | 4 ++-- .../skills/torchnpu-device-memory-lifecycle.json | 4 ++-- .../torchnpu-pinned-host-memory-lifecycle.json | 4 ++-- .../common/skills/assess-serving-preemption.json | 4 ++-- .../skills/assess-split-backward-schedule.json | 4 ++-- .../common/skills/audit-fixed-workload-claim.json | 4 ++-- .../skills/audit-h3-metal-fasth3-semantics.json | 4 ++-- .../common/skills/audit-h3-runtime-residency.json | 4 ++-- .../common/skills/audit-llama-runtime-state.json | 4 ++-- .../skills/audit-recurrent-kv-finalization.json | 4 ++-- .../common/skills/audit-sd-cpp-residency.json | 4 ++-- .../skills/audit-sglang-diffusion-lifetimes.json | 4 ++-- .../common/skills/audit-vllm-kv-lifetimes.json | 4 ++-- .../skills/bound-windowed-attention-replay.json | 4 ++-- .../skills/budget-speculative-verification.json | 4 ++-- .../common/skills/calibrate-serving-simulation.json | 4 ++-- .../common/skills/cgroup-resource-controls.json | 4 ++-- .../skills/checkpoint-recovery-consistency.json | 4 ++-- .../common/skills/choose-parallel-training-plan.json | 4 ++-- .../common/skills/cite-check-downstream-pr.json | 4 ++-- .../knowledge/common/skills/claim-contract.json | 4 ++-- .../inferrsi/knowledge/common/skills/comfy-ui.json | 4 ++-- .../knowledge/common/skills/competitor-scan.json | 4 ++-- .../common/skills/dcp-async-save-performance.json | 4 ++-- .../common/skills/dense-performance-evidence.json | 4 ++-- .../skills/diagnose-vllm-admission-pressure.json | 4 ++-- .../common/skills/diffusers-invocation-state.json | 4 ++-- .../common/skills/draft-acceptance-issue.json | 4 ++-- .../inferrsi/knowledge/common/skills/dwarfstar.json | 4 ++-- .../skills/evaluate-elastic-serving-placement.json | 4 ++-- .../common/skills/generation-cache-lifecycle.json | 4 ++-- .../common/skills/host-memory-management.json | 4 ++-- .../common/skills/host-overhead-optimization.json | 4 ++-- .../inferrsi/knowledge/common/skills/husky.json | 4 ++-- .../common/skills/inference-benchmark-design.json | 4 ++-- .../skills/inference-optimization-casebook.json | 12 ++++++------ .../common/skills/kernel-writing-guide.json | 4 ++-- .../knowledge/common/skills/kv-tier-management.json | 4 ++-- .../skills/lazy-evaluation-stream-lifetimes.json | 4 ++-- .../inferrsi/knowledge/common/skills/lily.json | 4 ++-- .../knowledge/common/skills/lithos-metal.json | 4 ++-- .../knowledge/common/skills/local-dream.json | 4 ++-- .../inferrsi/knowledge/common/skills/mistral-rs.json | 4 ++-- .../inferrsi/knowledge/common/skills/mlc-llm.json | 4 ++-- .../knowledge/common/skills/mlx-swift-lm.json | 4 ++-- .../knowledge/common/skills/npu-skill-factory.json | 4 ++-- .../common/skills/optimization-search-policy.json | 4 ++-- .../common/skills/os-process-placement.json | 4 ++-- .../common/skills/performance-ceiling-analysis.json | 4 ++-- .../common/skills/plan-activation-residency.json | 4 ++-- .../knowledge/common/skills/project-maya.json | 4 ++-- .../inferrsi/knowledge/common/skills/pytorch.json | 4 ++-- .../common/skills/repository-evolution-analysis.json | 4 ++-- .../knowledge/common/skills/rsi-evidence.json | 4 ++-- .../common/skills/skill-evolution-playbook.json | 4 ++-- .../inferrsi/knowledge/common/skills/splash.json | 4 ++-- .../common/skills/storage-io-measurement.json | 4 ++-- .../common/skills/training-input-pipeline.json | 4 ++-- .../common/skills/training-resource-budget.json | 4 ++-- .../common/skills/video-diffusion-optimization.json | 4 ++-- .../inferrsi/knowledge/common/skills/vllm-metal.json | 4 ++-- .../common/skills/weight-loading-optimization.json | 4 ++-- .../cuda/skills/audit-afd-p2p-exchanges.json | 4 ++-- .../cuda/skills/audit-flare-session-lifecycle.json | 4 ++-- .../cuda/skills/diagnose-attention-bottlenecks.json | 4 ++-- .../cuda/skills/host-overhead-profiling.json | 4 ++-- .../knowledge/cuda/skills/reason-hopper-kernels.json | 4 ++-- 81 files changed, 165 insertions(+), 165 deletions(-) diff --git a/docs/skill-coverage.md b/docs/skill-coverage.md index 680cb11cf..82414b456 100644 --- a/docs/skill-coverage.md +++ b/docs/skill-coverage.md @@ -4,7 +4,7 @@ Generated from the registry, the skill sidecars and the two maintained matrices. **Do not edit by hand**; regenerate after any registry, skill, matrix or frozen-plan change. -- Input digest: `sha256:e36cf2e6e982084dbf4e52014a6342a61bc1b49151d52ac0bf03f8e7ea975cff` +- Input digest: `sha256:d08d980c42000d245484033443c1bec3c6f6a3e2da5942400b80a1b3d49379fe` - Regenerate: `python tools/build_skill_coverage.py` - Verify without writing: `python tools/build_skill_coverage.py --check` diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/README.md b/src/inferrsi/data/repo_packs/inferrsi/knowledge/README.md index 06d738915..680d43ba7 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/README.md +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/README.md @@ -15,7 +15,7 @@ These are knowledge abstracts (`--kind knowledge`), with namespaced package IDs distinct from explicitly selected `checkout_ref` skills. Ascend/CUDA procedures use their corresponding backend; shared methods use `common`. -Source snapshot: [8115da15b9ef34978de500b08aa9f14ea790c35b](https://github.com/InferMatrixLab/InferRSI/tree/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library). +Source snapshot: [405033073d1140084da654c3701ac32cc4d1e4a1](https://github.com/InferMatrixLab/InferRSI/tree/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library). ```bash inferrsi knowledge search 'profiling' --repo-type inferrsi --limit 50 @@ -29,7 +29,7 @@ contain the current input paths and bytes (Git checkout filters are respected). ```bash python tools/build_skill_index.py --source-revision -python tools/build_skill_index.py --source-revision 8115da15b9ef34978de500b08aa9f14ea790c35b --check +python tools/build_skill_index.py --source-revision 405033073d1140084da654c3701ac32cc4d1e4a1 --check ``` `--check` detects missing, stale and unexpected generated records without writing. diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-aclgraph-capture-replay.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-aclgraph-capture-replay.json index ee9392c0b..3ea3e036d 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-aclgraph-capture-replay.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-aclgraph-capture-replay.json @@ -21,8 +21,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/hardware/ascend/ascend-aclgraph-capture-replay/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/hardware/ascend/ascend-aclgraph-capture-replay/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/hardware/ascend/ascend-aclgraph-capture-replay/SKILL.md", "src/inferrsi/data/skill_library/hardware/ascend/ascend-aclgraph-capture-replay/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-device-health.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-device-health.json index 70b46caa1..6fb347c13 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-device-health.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-device-health.json @@ -23,8 +23,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/hardware/ascend/ascend-device-health/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/hardware/ascend/ascend-device-health/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/hardware/ascend/ascend-device-health/SKILL.md", "src/inferrsi/data/skill_library/hardware/ascend/ascend-device-health/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-distributed-parallel.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-distributed-parallel.json index f3a2f5092..83acbe64c 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-distributed-parallel.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-distributed-parallel.json @@ -27,8 +27,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/hardware/ascend/ascend-distributed-parallel/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/hardware/ascend/ascend-distributed-parallel/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/hardware/ascend/ascend-distributed-parallel/SKILL.md", "src/inferrsi/data/skill_library/hardware/ascend/ascend-distributed-parallel/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-graph-mode.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-graph-mode.json index 6f7dedbe0..e54939a74 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-graph-mode.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-graph-mode.json @@ -26,8 +26,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/hardware/ascend/ascend-graph-mode/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/hardware/ascend/ascend-graph-mode/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/hardware/ascend/ascend-graph-mode/SKILL.md", "src/inferrsi/data/skill_library/hardware/ascend/ascend-graph-mode/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-model-adaptation.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-model-adaptation.json index b2970a363..d918c66e7 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-model-adaptation.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-model-adaptation.json @@ -22,8 +22,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/hardware/ascend/ascend-model-adaptation/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/hardware/ascend/ascend-model-adaptation/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/hardware/ascend/ascend-model-adaptation/SKILL.md", "src/inferrsi/data/skill_library/hardware/ascend/ascend-model-adaptation/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-operator-profiling.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-operator-profiling.json index 025c78bc9..1e4a12a9b 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-operator-profiling.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-operator-profiling.json @@ -23,8 +23,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/hardware/ascend/ascend-operator-profiling/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/hardware/ascend/ascend-operator-profiling/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/hardware/ascend/ascend-operator-profiling/SKILL.md", "src/inferrsi/data/skill_library/hardware/ascend/ascend-operator-profiling/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-operator-tuning.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-operator-tuning.json index b39027a2c..4febffeb1 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-operator-tuning.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-operator-tuning.json @@ -24,8 +24,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/hardware/ascend/ascend-operator-tuning/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/hardware/ascend/ascend-operator-tuning/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/hardware/ascend/ascend-operator-tuning/SKILL.md", "src/inferrsi/data/skill_library/hardware/ascend/ascend-operator-tuning/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-optimization-casebook.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-optimization-casebook.json index 37956bcf2..14243ca92 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-optimization-casebook.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-optimization-casebook.json @@ -22,8 +22,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/hardware/ascend/ascend-optimization-casebook/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/hardware/ascend/ascend-optimization-casebook/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/hardware/ascend/ascend-optimization-casebook/SKILL.md", "src/inferrsi/data/skill_library/hardware/ascend/ascend-optimization-casebook/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-performance-analysis.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-performance-analysis.json index 18fbb990e..3da7d2ea4 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-performance-analysis.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-performance-analysis.json @@ -22,8 +22,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/hardware/ascend/ascend-performance-analysis/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/hardware/ascend/ascend-performance-analysis/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/hardware/ascend/ascend-performance-analysis/SKILL.md", "src/inferrsi/data/skill_library/hardware/ascend/ascend-performance-analysis/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-video-diffusion-optimization.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-video-diffusion-optimization.json index 54dd5bc6f..1ad0e06f7 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-video-diffusion-optimization.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-video-diffusion-optimization.json @@ -26,8 +26,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/hardware/ascend/ascend-video-diffusion-optimization/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/hardware/ascend/ascend-video-diffusion-optimization/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/hardware/ascend/ascend-video-diffusion-optimization/SKILL.md", "src/inferrsi/data/skill_library/hardware/ascend/ascend-video-diffusion-optimization/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascendc-tiling-design.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascendc-tiling-design.json index 0de8de571..27ac8b814 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascendc-tiling-design.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascendc-tiling-design.json @@ -24,8 +24,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/hardware/ascend/ascendc-tiling-design/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/hardware/ascend/ascendc-tiling-design/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/hardware/ascend/ascendc-tiling-design/SKILL.md", "src/inferrsi/data/skill_library/hardware/ascend/ascendc-tiling-design/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/cann-attribution-overlap.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/cann-attribution-overlap.json index fba89f6ca..f46d8f1b5 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/cann-attribution-overlap.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/cann-attribution-overlap.json @@ -24,8 +24,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/hardware/ascend/cann-attribution-overlap/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/hardware/ascend/cann-attribution-overlap/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/hardware/ascend/cann-attribution-overlap/SKILL.md", "src/inferrsi/data/skill_library/hardware/ascend/cann-attribution-overlap/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/torch-npu.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/torch-npu.json index deb5a8d02..2a0ec342c 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/torch-npu.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/torch-npu.json @@ -20,8 +20,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/frameworks/torch-npu/torch-npu/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/frameworks/torch-npu/torch-npu/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/frameworks/torch-npu/torch-npu/SKILL.md", "src/inferrsi/data/skill_library/frameworks/torch-npu/torch-npu/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/torchnpu-device-memory-lifecycle.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/torchnpu-device-memory-lifecycle.json index cf82999ed..8e389e825 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/torchnpu-device-memory-lifecycle.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/torchnpu-device-memory-lifecycle.json @@ -22,8 +22,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/hardware/ascend/torchnpu-device-memory-lifecycle/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/hardware/ascend/torchnpu-device-memory-lifecycle/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/hardware/ascend/torchnpu-device-memory-lifecycle/SKILL.md", "src/inferrsi/data/skill_library/hardware/ascend/torchnpu-device-memory-lifecycle/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/torchnpu-pinned-host-memory-lifecycle.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/torchnpu-pinned-host-memory-lifecycle.json index 0da5a9009..3f7d26e1a 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/torchnpu-pinned-host-memory-lifecycle.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/torchnpu-pinned-host-memory-lifecycle.json @@ -22,8 +22,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/hardware/ascend/torchnpu-pinned-host-memory-lifecycle/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/hardware/ascend/torchnpu-pinned-host-memory-lifecycle/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/hardware/ascend/torchnpu-pinned-host-memory-lifecycle/SKILL.md", "src/inferrsi/data/skill_library/hardware/ascend/torchnpu-pinned-host-memory-lifecycle/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/assess-serving-preemption.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/assess-serving-preemption.json index 4b0644d21..dbc123f59 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/assess-serving-preemption.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/assess-serving-preemption.json @@ -16,8 +16,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/methods/assess-serving-preemption/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/methods/assess-serving-preemption/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/methods/assess-serving-preemption/SKILL.md", "src/inferrsi/data/skill_library/methods/assess-serving-preemption/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/assess-split-backward-schedule.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/assess-split-backward-schedule.json index 7f38c7cb2..7c34f2658 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/assess-split-backward-schedule.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/assess-split-backward-schedule.json @@ -16,8 +16,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/methods/assess-split-backward-schedule/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/methods/assess-split-backward-schedule/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/methods/assess-split-backward-schedule/SKILL.md", "src/inferrsi/data/skill_library/methods/assess-split-backward-schedule/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-fixed-workload-claim.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-fixed-workload-claim.json index f4c665268..e2dc473a2 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-fixed-workload-claim.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-fixed-workload-claim.json @@ -20,8 +20,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/methods/audit-fixed-workload-claim/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/methods/audit-fixed-workload-claim/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/methods/audit-fixed-workload-claim/SKILL.md", "src/inferrsi/data/skill_library/methods/audit-fixed-workload-claim/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-h3-metal-fasth3-semantics.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-h3-metal-fasth3-semantics.json index bb7e7f3a5..de43d59d4 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-h3-metal-fasth3-semantics.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-h3-metal-fasth3-semantics.json @@ -29,8 +29,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/h3-metal/audit-h3-metal-fasth3-semantics/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/h3-metal/audit-h3-metal-fasth3-semantics/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/h3-metal/audit-h3-metal-fasth3-semantics/LICENSE", "src/inferrsi/data/skill_library/repositories/h3-metal/audit-h3-metal-fasth3-semantics/SKILL.md", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-h3-runtime-residency.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-h3-runtime-residency.json index 238d0d0b9..19ba9a1bc 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-h3-runtime-residency.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-h3-runtime-residency.json @@ -30,8 +30,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/h3-c/audit-h3-runtime-residency/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/h3-c/audit-h3-runtime-residency/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/h3-c/audit-h3-runtime-residency/LICENSE", "src/inferrsi/data/skill_library/repositories/h3-c/audit-h3-runtime-residency/SKILL.md", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-llama-runtime-state.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-llama-runtime-state.json index 8efdd50cd..d35fc3b5f 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-llama-runtime-state.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-llama-runtime-state.json @@ -23,8 +23,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/llama-cpp/audit-llama-runtime-state/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/llama-cpp/audit-llama-runtime-state/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/llama-cpp/audit-llama-runtime-state/SKILL.md", "src/inferrsi/data/skill_library/repositories/llama-cpp/audit-llama-runtime-state/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-recurrent-kv-finalization.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-recurrent-kv-finalization.json index f3e33ac15..3d14bc813 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-recurrent-kv-finalization.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-recurrent-kv-finalization.json @@ -26,8 +26,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/vllm-rlt/audit-recurrent-kv-finalization/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/vllm-rlt/audit-recurrent-kv-finalization/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/vllm-rlt/audit-recurrent-kv-finalization/LICENSE", "src/inferrsi/data/skill_library/repositories/vllm-rlt/audit-recurrent-kv-finalization/SKILL.md", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-sd-cpp-residency.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-sd-cpp-residency.json index 9a398fafc..97a69dc16 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-sd-cpp-residency.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-sd-cpp-residency.json @@ -25,8 +25,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/stable-diffusion-cpp/audit-sd-cpp-residency/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/stable-diffusion-cpp/audit-sd-cpp-residency/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/stable-diffusion-cpp/audit-sd-cpp-residency/LICENSE", "src/inferrsi/data/skill_library/repositories/stable-diffusion-cpp/audit-sd-cpp-residency/SKILL.md", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-sglang-diffusion-lifetimes.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-sglang-diffusion-lifetimes.json index da6a69568..3b698c53b 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-sglang-diffusion-lifetimes.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-sglang-diffusion-lifetimes.json @@ -23,8 +23,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/sglang/audit-sglang-diffusion-lifetimes/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/sglang/audit-sglang-diffusion-lifetimes/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/sglang/audit-sglang-diffusion-lifetimes/LICENSE", "src/inferrsi/data/skill_library/repositories/sglang/audit-sglang-diffusion-lifetimes/SKILL.md", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-vllm-kv-lifetimes.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-vllm-kv-lifetimes.json index f3873bf51..c6116576c 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-vllm-kv-lifetimes.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-vllm-kv-lifetimes.json @@ -21,8 +21,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/vllm/audit-vllm-kv-lifetimes/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/vllm/audit-vllm-kv-lifetimes/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/vllm/audit-vllm-kv-lifetimes/SKILL.md", "src/inferrsi/data/skill_library/repositories/vllm/audit-vllm-kv-lifetimes/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/bound-windowed-attention-replay.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/bound-windowed-attention-replay.json index 5f5885cf9..a9feb1339 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/bound-windowed-attention-replay.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/bound-windowed-attention-replay.json @@ -23,8 +23,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/deepinfer/bound-windowed-attention-replay/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/deepinfer/bound-windowed-attention-replay/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/deepinfer/bound-windowed-attention-replay/SKILL.md", "src/inferrsi/data/skill_library/repositories/deepinfer/bound-windowed-attention-replay/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/budget-speculative-verification.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/budget-speculative-verification.json index a2e12f98e..c40ebc02a 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/budget-speculative-verification.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/budget-speculative-verification.json @@ -23,8 +23,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/deepinfer/budget-speculative-verification/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/deepinfer/budget-speculative-verification/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/deepinfer/budget-speculative-verification/SKILL.md", "src/inferrsi/data/skill_library/repositories/deepinfer/budget-speculative-verification/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/calibrate-serving-simulation.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/calibrate-serving-simulation.json index 4c0129585..036d45dcd 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/calibrate-serving-simulation.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/calibrate-serving-simulation.json @@ -16,8 +16,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/methods/calibrate-serving-simulation/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/methods/calibrate-serving-simulation/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/methods/calibrate-serving-simulation/SKILL.md", "src/inferrsi/data/skill_library/methods/calibrate-serving-simulation/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/cgroup-resource-controls.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/cgroup-resource-controls.json index 57466f972..ac4952723 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/cgroup-resource-controls.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/cgroup-resource-controls.json @@ -25,8 +25,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/hardware/host/cgroup-resource-controls/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/hardware/host/cgroup-resource-controls/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/hardware/host/cgroup-resource-controls/SKILL.md", "src/inferrsi/data/skill_library/hardware/host/cgroup-resource-controls/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/checkpoint-recovery-consistency.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/checkpoint-recovery-consistency.json index 0f98dcd89..ccf4032dd 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/checkpoint-recovery-consistency.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/checkpoint-recovery-consistency.json @@ -22,8 +22,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/methods/checkpoint-recovery-consistency/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/methods/checkpoint-recovery-consistency/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/methods/checkpoint-recovery-consistency/SKILL.md", "src/inferrsi/data/skill_library/methods/checkpoint-recovery-consistency/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/choose-parallel-training-plan.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/choose-parallel-training-plan.json index d0b8f03f3..2b2781f9e 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/choose-parallel-training-plan.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/choose-parallel-training-plan.json @@ -16,8 +16,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/methods/choose-parallel-training-plan/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/methods/choose-parallel-training-plan/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/methods/choose-parallel-training-plan/SKILL.md", "src/inferrsi/data/skill_library/methods/choose-parallel-training-plan/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/cite-check-downstream-pr.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/cite-check-downstream-pr.json index db7d7a4e0..df380558a 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/cite-check-downstream-pr.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/cite-check-downstream-pr.json @@ -22,8 +22,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/workflows/cite-check-downstream-pr/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/workflows/cite-check-downstream-pr/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/skill-index.json", "src/inferrsi/data/skill_library/workflows/cite-check-downstream-pr/SKILL.md", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/claim-contract.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/claim-contract.json index b9c760916..b4fa961da 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/claim-contract.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/claim-contract.json @@ -22,8 +22,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/workflows/claim-contract/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/workflows/claim-contract/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/skill-index.json", "src/inferrsi/data/skill_library/workflows/claim-contract/SKILL.md", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/comfy-ui.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/comfy-ui.json index fb37a8146..a4dd5ba8c 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/comfy-ui.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/comfy-ui.json @@ -22,8 +22,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/comfy-ui/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/comfy-ui/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/comfy-ui/LICENSE", "src/inferrsi/data/skill_library/repositories/comfy-ui/SKILL.md", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/competitor-scan.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/competitor-scan.json index 6b9978048..abc9a3d0e 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/competitor-scan.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/competitor-scan.json @@ -21,8 +21,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/workflows/competitor-scan/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/workflows/competitor-scan/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/skill-index.json", "src/inferrsi/data/skill_library/workflows/competitor-scan/SKILL.md", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/dcp-async-save-performance.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/dcp-async-save-performance.json index 3f35fb7c5..515cbe794 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/dcp-async-save-performance.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/dcp-async-save-performance.json @@ -23,8 +23,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/frameworks/pytorch/dcp-async-save-performance/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/frameworks/pytorch/dcp-async-save-performance/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/frameworks/pytorch/dcp-async-save-performance/SKILL.md", "src/inferrsi/data/skill_library/frameworks/pytorch/dcp-async-save-performance/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/dense-performance-evidence.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/dense-performance-evidence.json index fa36bc454..609dd3d71 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/dense-performance-evidence.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/dense-performance-evidence.json @@ -21,8 +21,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/methods/dense-performance-evidence/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/methods/dense-performance-evidence/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/methods/dense-performance-evidence/SKILL.md", "src/inferrsi/data/skill_library/methods/dense-performance-evidence/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/diagnose-vllm-admission-pressure.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/diagnose-vllm-admission-pressure.json index 6b458bdbc..72009c0e0 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/diagnose-vllm-admission-pressure.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/diagnose-vllm-admission-pressure.json @@ -21,8 +21,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/vllm/diagnose-vllm-admission-pressure/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/vllm/diagnose-vllm-admission-pressure/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/vllm/diagnose-vllm-admission-pressure/SKILL.md", "src/inferrsi/data/skill_library/repositories/vllm/diagnose-vllm-admission-pressure/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/diffusers-invocation-state.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/diffusers-invocation-state.json index 6f367301f..0f196d409 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/diffusers-invocation-state.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/diffusers-invocation-state.json @@ -22,8 +22,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/frameworks/diffusers/diffusers-invocation-state/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/frameworks/diffusers/diffusers-invocation-state/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/frameworks/diffusers/diffusers-invocation-state/SKILL.md", "src/inferrsi/data/skill_library/frameworks/diffusers/diffusers-invocation-state/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/draft-acceptance-issue.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/draft-acceptance-issue.json index a6b8f26ef..dbdfffbb5 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/draft-acceptance-issue.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/draft-acceptance-issue.json @@ -22,8 +22,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/workflows/draft-acceptance-issue/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/workflows/draft-acceptance-issue/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/skill-index.json", "src/inferrsi/data/skill_library/workflows/draft-acceptance-issue/SKILL.md", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/dwarfstar.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/dwarfstar.json index e8307d3f3..eda8c10b5 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/dwarfstar.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/dwarfstar.json @@ -25,8 +25,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/dwarfstar/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/dwarfstar/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/dwarfstar/LICENSE", "src/inferrsi/data/skill_library/repositories/dwarfstar/SKILL.md", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/evaluate-elastic-serving-placement.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/evaluate-elastic-serving-placement.json index a412750b8..ab4919382 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/evaluate-elastic-serving-placement.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/evaluate-elastic-serving-placement.json @@ -16,8 +16,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/methods/evaluate-elastic-serving-placement/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/methods/evaluate-elastic-serving-placement/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/methods/evaluate-elastic-serving-placement/SKILL.md", "src/inferrsi/data/skill_library/methods/evaluate-elastic-serving-placement/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/generation-cache-lifecycle.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/generation-cache-lifecycle.json index 49c108bcc..eeb0bafc4 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/generation-cache-lifecycle.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/generation-cache-lifecycle.json @@ -22,8 +22,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/mlx-lm/generation-cache-lifecycle/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/mlx-lm/generation-cache-lifecycle/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/mlx-lm/generation-cache-lifecycle/SKILL.md", "src/inferrsi/data/skill_library/repositories/mlx-lm/generation-cache-lifecycle/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/host-memory-management.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/host-memory-management.json index f15b08ef6..5f6d1b032 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/host-memory-management.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/host-memory-management.json @@ -30,8 +30,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/hardware/host/host-memory-management/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/hardware/host/host-memory-management/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/hardware/host/host-memory-management/SKILL.md", "src/inferrsi/data/skill_library/hardware/host/host-memory-management/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/host-overhead-optimization.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/host-overhead-optimization.json index e7c2ea831..0bde0acd9 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/host-overhead-optimization.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/host-overhead-optimization.json @@ -22,8 +22,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/hardware/host/host-overhead-optimization/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/hardware/host/host-overhead-optimization/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/hardware/host/host-overhead-optimization/SKILL.md", "src/inferrsi/data/skill_library/hardware/host/host-overhead-optimization/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/husky.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/husky.json index d6ecd221c..6b94a04f7 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/husky.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/husky.json @@ -21,8 +21,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/husky/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/husky/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/husky/SKILL.md", "src/inferrsi/data/skill_library/repositories/husky/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/inference-benchmark-design.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/inference-benchmark-design.json index b671bc29f..9f74f0688 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/inference-benchmark-design.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/inference-benchmark-design.json @@ -35,8 +35,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/methods/inference-benchmark-design/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/methods/inference-benchmark-design/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/methods/inference-benchmark-design/SKILL.md", "src/inferrsi/data/skill_library/methods/inference-benchmark-design/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/inference-optimization-casebook.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/inference-optimization-casebook.json index 384974e45..15372785f 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/inference-optimization-casebook.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/inference-optimization-casebook.json @@ -21,8 +21,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/SKILL.md", "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/admission-drain-occupancy.json", @@ -42,7 +42,7 @@ "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/metadata.json", "src/inferrsi/data/skill_library/skill-index.json" ], - "content_sha256": "e6ff2de3835ad455df1c5bab7ad002d8489d8c7ef9abeab9660206f602cc8fe7" + "content_sha256": "f146497195d3c33a9a4a57488791b0e82d9d7122d0f0fe3a506a484c46c38286" } ], "index_type": "packaged-skill", @@ -51,7 +51,7 @@ }, "checkout_ref": "checkout:inference-optimization-casebook", "source_files": { - "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/SKILL.md": "e6ff2de3835ad455df1c5bab7ad002d8489d8c7ef9abeab9660206f602cc8fe7", + "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/SKILL.md": "f146497195d3c33a9a4a57488791b0e82d9d7122d0f0fe3a506a484c46c38286", "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/admission-drain-occupancy.json": "3319e23bbe2bd40e2d45fb89dc87bdc92d4df2b5ddd8086fcf84813b43885b7c", "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/admission-drain-occupancy.md": "e88bbac5d253a8e7e3a3d5e21047f23f2970c11ac23887d4303ab94fca52dd27", "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/drama3-rejected-directions.json": "2cc0a526fe0509551607a03e7626dd3d7844e643c4afb194837f75cddead8b12", @@ -62,8 +62,8 @@ "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen-b300-rejected-directions.md": "cce2930f544a0f78f21797fba4d8ded298e4a81f76c34d7977b0e31d361bc738", "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.json": "7d02a7435560745d642dc1137859e2a56914327f1235c52e726b277fdf470a27", "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-cuda-graph-state-not-accumulating.md": "8a51ed2c55d0aebff87fddfba8eb9cfbb415e014f7534d3fe58cbd72e8f5deb5", - "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.json": "54b686f7a1f935ace148c49ee7ec343343dc38877fe9d5a6efadf46ef1a6e27e", - "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md": "2f53bb0559231cc43da592fc47065209143e9d85e5b67eb129126f53c777953f", + "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.json": "fcd07aa35d5729d4486d1e34ec38f648a199828fa54e8d5261c490d9df56f27f", + "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-silent-capacity-and-error-failures.md": "50cb3fdf75415ec260ed8d6573a4a423af7c3ff6ac1e0dae2cce21ac873344f0", "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.json": "07fbd0ff59a1fff168f48b7e8084f038bd855f9cb493913acc850bf7d492abd0", "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/qwen38-upstream-fp8-kernels-vs-own-triton.md": "7b01b9719265430ed2be83e0535def09af5393035c714041fb16e6387ed0d6df", "src/inferrsi/data/skill_library/methods/inference-optimization-casebook/metadata.json": "963488658f327d2379183d70357eabf832d70a6a70db2aa67f548493f4088d97", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/kernel-writing-guide.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/kernel-writing-guide.json index d4d9423e2..bb2587900 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/kernel-writing-guide.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/kernel-writing-guide.json @@ -39,8 +39,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/methods/kernel-writing-guide/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/methods/kernel-writing-guide/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/methods/kernel-writing-guide/SKILL.md", "src/inferrsi/data/skill_library/methods/kernel-writing-guide/cases/causal-conv1d-dispatch-state.md", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/kv-tier-management.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/kv-tier-management.json index 5bbd97561..475693303 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/kv-tier-management.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/kv-tier-management.json @@ -24,8 +24,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/lmcache/kv-tier-management/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/lmcache/kv-tier-management/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/lmcache/kv-tier-management/SKILL.md", "src/inferrsi/data/skill_library/repositories/lmcache/kv-tier-management/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/lazy-evaluation-stream-lifetimes.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/lazy-evaluation-stream-lifetimes.json index 76d071a22..e27788a13 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/lazy-evaluation-stream-lifetimes.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/lazy-evaluation-stream-lifetimes.json @@ -22,8 +22,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/frameworks/mlx/lazy-evaluation-stream-lifetimes/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/frameworks/mlx/lazy-evaluation-stream-lifetimes/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/frameworks/mlx/lazy-evaluation-stream-lifetimes/SKILL.md", "src/inferrsi/data/skill_library/frameworks/mlx/lazy-evaluation-stream-lifetimes/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/lily.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/lily.json index cfb60ddca..18cae151a 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/lily.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/lily.json @@ -25,8 +25,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/lily/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/lily/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/lily/LICENSE", "src/inferrsi/data/skill_library/repositories/lily/NOTICE.md", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/lithos-metal.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/lithos-metal.json index 2cf62297d..ce3437a68 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/lithos-metal.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/lithos-metal.json @@ -21,8 +21,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/lithos-metal/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/lithos-metal/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/lithos-metal/LICENSE", "src/inferrsi/data/skill_library/repositories/lithos-metal/SKILL.md", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/local-dream.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/local-dream.json index 18e4aa3bf..28f7560ec 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/local-dream.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/local-dream.json @@ -29,8 +29,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/local-dream/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/local-dream/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/local-dream/SKILL.md", "src/inferrsi/data/skill_library/repositories/local-dream/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/mistral-rs.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/mistral-rs.json index 592dd3812..a70f24b5e 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/mistral-rs.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/mistral-rs.json @@ -26,8 +26,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/mistral-rs/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/mistral-rs/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/mistral-rs/LICENSE", "src/inferrsi/data/skill_library/repositories/mistral-rs/SKILL.md", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/mlc-llm.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/mlc-llm.json index 860b73efc..c785b69d1 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/mlc-llm.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/mlc-llm.json @@ -27,8 +27,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/mlc-llm/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/mlc-llm/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/mlc-llm/LICENSE", "src/inferrsi/data/skill_library/repositories/mlc-llm/SKILL.md", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/mlx-swift-lm.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/mlx-swift-lm.json index b5a9f7d91..72e94695f 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/mlx-swift-lm.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/mlx-swift-lm.json @@ -27,8 +27,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/mlx-swift-lm/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/mlx-swift-lm/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/mlx-swift-lm/LICENSE", "src/inferrsi/data/skill_library/repositories/mlx-swift-lm/SKILL.md", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/npu-skill-factory.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/npu-skill-factory.json index 0b6fbb8ad..ce84febd3 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/npu-skill-factory.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/npu-skill-factory.json @@ -22,8 +22,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/workflows/npu-skill-factory/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/workflows/npu-skill-factory/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/skill-index.json", "src/inferrsi/data/skill_library/workflows/npu-skill-factory/SKILL.md", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/optimization-search-policy.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/optimization-search-policy.json index e4f81d8d1..1d01f2d03 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/optimization-search-policy.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/optimization-search-policy.json @@ -21,8 +21,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/methods/optimization-search-policy/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/methods/optimization-search-policy/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/methods/optimization-search-policy/SKILL.md", "src/inferrsi/data/skill_library/methods/optimization-search-policy/cases/sarathi-chunk-budget.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/os-process-placement.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/os-process-placement.json index 36f83e9db..feeaf936d 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/os-process-placement.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/os-process-placement.json @@ -24,8 +24,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/hardware/host/os-process-placement/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/hardware/host/os-process-placement/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/hardware/host/os-process-placement/SKILL.md", "src/inferrsi/data/skill_library/hardware/host/os-process-placement/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/performance-ceiling-analysis.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/performance-ceiling-analysis.json index 683609a41..863da0ebd 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/performance-ceiling-analysis.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/performance-ceiling-analysis.json @@ -22,8 +22,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/methods/performance-ceiling-analysis/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/methods/performance-ceiling-analysis/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/methods/performance-ceiling-analysis/SKILL.md", "src/inferrsi/data/skill_library/methods/performance-ceiling-analysis/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/plan-activation-residency.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/plan-activation-residency.json index 23d3810b7..25e1140c8 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/plan-activation-residency.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/plan-activation-residency.json @@ -16,8 +16,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/methods/plan-activation-residency/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/methods/plan-activation-residency/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/methods/plan-activation-residency/SKILL.md", "src/inferrsi/data/skill_library/methods/plan-activation-residency/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/project-maya.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/project-maya.json index 17c8095d5..609fca2fb 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/project-maya.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/project-maya.json @@ -25,8 +25,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/project-maya/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/project-maya/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/project-maya/LICENSE", "src/inferrsi/data/skill_library/repositories/project-maya/SKILL.md", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/pytorch.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/pytorch.json index 09306d843..18b0523a8 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/pytorch.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/pytorch.json @@ -24,8 +24,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/frameworks/pytorch/pytorch/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/frameworks/pytorch/pytorch/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/frameworks/pytorch/pytorch/LICENSE", "src/inferrsi/data/skill_library/frameworks/pytorch/pytorch/SKILL.md", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/repository-evolution-analysis.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/repository-evolution-analysis.json index 87ac2caa3..a85667e43 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/repository-evolution-analysis.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/repository-evolution-analysis.json @@ -21,8 +21,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/methods/repository-evolution-analysis/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/methods/repository-evolution-analysis/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/methods/repository-evolution-analysis/SKILL.md", "src/inferrsi/data/skill_library/methods/repository-evolution-analysis/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/rsi-evidence.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/rsi-evidence.json index 3a66d4cde..e5397c9c4 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/rsi-evidence.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/rsi-evidence.json @@ -22,8 +22,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/workflows/rsi-evidence/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/workflows/rsi-evidence/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/skill-index.json", "src/inferrsi/data/skill_library/workflows/rsi-evidence/SKILL.md", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/skill-evolution-playbook.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/skill-evolution-playbook.json index f360339fa..72dd8ecbf 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/skill-evolution-playbook.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/skill-evolution-playbook.json @@ -21,8 +21,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/methods/skill-evolution-playbook/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/methods/skill-evolution-playbook/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/methods/skill-evolution-playbook/SKILL.md", "src/inferrsi/data/skill_library/methods/skill-evolution-playbook/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/splash.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/splash.json index 23c775be0..62e049575 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/splash.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/splash.json @@ -25,8 +25,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/splash/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/splash/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/splash/LICENSE", "src/inferrsi/data/skill_library/repositories/splash/SKILL.md", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/storage-io-measurement.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/storage-io-measurement.json index 130c9cf61..ed39ce04f 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/storage-io-measurement.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/storage-io-measurement.json @@ -24,8 +24,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/hardware/host/storage-io-measurement/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/hardware/host/storage-io-measurement/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/hardware/host/storage-io-measurement/SKILL.md", "src/inferrsi/data/skill_library/hardware/host/storage-io-measurement/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/training-input-pipeline.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/training-input-pipeline.json index 25f5955b9..0e6def3f7 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/training-input-pipeline.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/training-input-pipeline.json @@ -23,8 +23,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/methods/training-input-pipeline/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/methods/training-input-pipeline/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/methods/training-input-pipeline/SKILL.md", "src/inferrsi/data/skill_library/methods/training-input-pipeline/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/training-resource-budget.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/training-resource-budget.json index 5027f5a98..778d32d37 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/training-resource-budget.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/training-resource-budget.json @@ -23,8 +23,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/methods/training-resource-budget/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/methods/training-resource-budget/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/methods/training-resource-budget/SKILL.md", "src/inferrsi/data/skill_library/methods/training-resource-budget/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/video-diffusion-optimization.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/video-diffusion-optimization.json index ba51c5a5f..e9b386a00 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/video-diffusion-optimization.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/video-diffusion-optimization.json @@ -26,8 +26,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/methods/video-diffusion-optimization/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/methods/video-diffusion-optimization/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/methods/video-diffusion-optimization/SKILL.md", "src/inferrsi/data/skill_library/methods/video-diffusion-optimization/assets/campaign.template.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/vllm-metal.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/vllm-metal.json index e953111bf..eeb38abec 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/vllm-metal.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/vllm-metal.json @@ -26,8 +26,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/vllm-metal/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/vllm-metal/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/vllm-metal/LICENSE", "src/inferrsi/data/skill_library/repositories/vllm-metal/SKILL.md", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/weight-loading-optimization.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/weight-loading-optimization.json index ecad06118..efe22dc04 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/weight-loading-optimization.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/weight-loading-optimization.json @@ -29,8 +29,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/methods/weight-loading-optimization/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/methods/weight-loading-optimization/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/methods/weight-loading-optimization/SKILL.md", "src/inferrsi/data/skill_library/methods/weight-loading-optimization/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/audit-afd-p2p-exchanges.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/audit-afd-p2p-exchanges.json index 686bf7f82..b63f8a5e8 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/audit-afd-p2p-exchanges.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/audit-afd-p2p-exchanges.json @@ -23,8 +23,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/afd-plugin/audit-afd-p2p-exchanges/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/afd-plugin/audit-afd-p2p-exchanges/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/afd-plugin/audit-afd-p2p-exchanges/SKILL.md", "src/inferrsi/data/skill_library/repositories/afd-plugin/audit-afd-p2p-exchanges/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/audit-flare-session-lifecycle.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/audit-flare-session-lifecycle.json index 8227a0045..b5b015207 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/audit-flare-session-lifecycle.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/audit-flare-session-lifecycle.json @@ -29,8 +29,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/repositories/flare-local-inference/audit-flare-session-lifecycle/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/repositories/flare-local-inference/audit-flare-session-lifecycle/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/repositories/flare-local-inference/audit-flare-session-lifecycle/LICENSE", "src/inferrsi/data/skill_library/repositories/flare-local-inference/audit-flare-session-lifecycle/SKILL.md", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/diagnose-attention-bottlenecks.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/diagnose-attention-bottlenecks.json index eb8eb6dba..2760b2cbc 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/diagnose-attention-bottlenecks.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/diagnose-attention-bottlenecks.json @@ -16,8 +16,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/hardware/cuda/diagnose-attention-bottlenecks/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/hardware/cuda/diagnose-attention-bottlenecks/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/hardware/cuda/diagnose-attention-bottlenecks/SKILL.md", "src/inferrsi/data/skill_library/hardware/cuda/diagnose-attention-bottlenecks/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/host-overhead-profiling.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/host-overhead-profiling.json index f79630e3d..2d659023d 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/host-overhead-profiling.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/host-overhead-profiling.json @@ -22,8 +22,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/hardware/cuda/host-overhead-profiling/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/hardware/cuda/host-overhead-profiling/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/hardware/cuda/host-overhead-profiling/SKILL.md", "src/inferrsi/data/skill_library/hardware/cuda/host-overhead-profiling/metadata.json", diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/reason-hopper-kernels.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/reason-hopper-kernels.json index 7a0c4d427..8c2f89fb4 100644 --- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/reason-hopper-kernels.json +++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/reason-hopper-kernels.json @@ -36,8 +36,8 @@ "sources": [ { "type": "git", - "url": "https://github.com/InferMatrixLab/InferRSI/blob/8115da15b9ef34978de500b08aa9f14ea790c35b/src/inferrsi/data/skill_library/hardware/cuda/reason-hopper-kernels/SKILL.md", - "commit": "8115da15b9ef34978de500b08aa9f14ea790c35b", + "url": "https://github.com/InferMatrixLab/InferRSI/blob/405033073d1140084da654c3701ac32cc4d1e4a1/src/inferrsi/data/skill_library/hardware/cuda/reason-hopper-kernels/SKILL.md", + "commit": "405033073d1140084da654c3701ac32cc4d1e4a1", "paths": [ "src/inferrsi/data/skill_library/hardware/cuda/reason-hopper-kernels/SKILL.md", "src/inferrsi/data/skill_library/hardware/cuda/reason-hopper-kernels/metadata.json",