AutomatosX commited on
Commit
a3a6694
·
verified ·
1 Parent(s): dadc544

Correct evidenced MXFP8 container mode and bind runtime format audit

Browse files
Files changed (4) hide show
  1. README.md +13 -0
  2. axquant_manifest.json +10 -5
  3. config.json +0 -0
  4. runtime_audit.json +62 -0
README.md CHANGED
@@ -19,6 +19,19 @@ tags:
19
  - sentence-similarity
20
  ---
21
 
 
 
 
 
 
 
 
 
 
 
 
 
 
22
  # AX-Qwen3-Embedding-0.6B-MLX-AXQ-MXFP8 — 8.25 BPW measured main
23
 
24
  An **AXQuant (AXQ)** mixed-precision MLX checkpoint for Apple Silicon, converted directly from
 
19
  - sentence-similarity
20
  ---
21
 
22
+ ## Runtime format audit (2026-10-06)
23
+
24
+ The quantization container mode was corrected from `affine` to `mxfp8`
25
+ in both config blocks after inspecting every quantized module header.
26
+ Weight bytes and per-module precision assignments are unchanged.
27
+ This family has no n-gram tensors; no n-gram file or declaration was added.
28
+ This remote format audit does not grant an oMLX/MTPLX runtime profile or a successful load/generation claim.
29
+
30
+ See [runtime_audit.json](runtime_audit.json) for pinned config/index/header
31
+ bindings, architecture, physical-format findings, and applied corrections.
32
+ This is development evidence; no quality, MTP exactness, speed, or certification
33
+ claim is added. Historical evidence stays bound to its original revision.
34
+
35
  # AX-Qwen3-Embedding-0.6B-MLX-AXQ-MXFP8 — 8.25 BPW measured main
36
 
37
  An **AXQuant (AXQ)** mixed-precision MLX checkpoint for Apple Silicon, converted directly from
axquant_manifest.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
  "axquant_version": "1.9.0",
3
  "calibration": null,
4
- "created_at": "2026-10-04T02:06:14.423219Z",
5
  "effective_bpw": 9.000770006857874,
6
  "files": [
7
  {
@@ -11,8 +11,8 @@
11
  },
12
  {
13
  "path": "README.md",
14
- "sha256": "14d1f9c305e01f756995a5398fcfa0bd8f4b372da7804ef0e12e8ed68ab451c7",
15
- "size_bytes": 7769
16
  },
17
  {
18
  "path": "axquant_plan.json",
@@ -41,8 +41,8 @@
41
  },
42
  {
43
  "path": "config.json",
44
- "sha256": "3aeb949fad423de8a2d7c0c6d9bb52baf18a1c92e72b5761040f31ff1a3eebd0",
45
- "size_bytes": 54577
46
  },
47
  {
48
  "path": "config_sentence_transformers.json",
@@ -74,6 +74,11 @@
74
  "sha256": "84e40c8e006c9b1d6c122e02cba9b02458120b5fb0c87b746c41e0207cf642cf",
75
  "size_bytes": 349
76
  },
 
 
 
 
 
77
  {
78
  "path": "tokenizer.json",
79
  "sha256": "93623af029cdc69b87f2864d3b2cc2424fdf16684f15e139b5b9d08ec34ced91",
 
1
  {
2
  "axquant_version": "1.9.0",
3
  "calibration": null,
4
+ "created_at": "2026-10-06T21:01:14.077738+00:00",
5
  "effective_bpw": 9.000770006857874,
6
  "files": [
7
  {
 
11
  },
12
  {
13
  "path": "README.md",
14
+ "sha256": "77b189e19bdfa84fd8e9df7172bb7e0b371991f7cf85c20b35752f21a27800e6",
15
+ "size_bytes": 8506
16
  },
17
  {
18
  "path": "axquant_plan.json",
 
41
  },
42
  {
43
  "path": "config.json",
44
+ "sha256": "b49450c5c373f46fa637822bbbd466ada245fb3b58c40e589d41a8932070670f",
45
+ "size_bytes": 44242
46
  },
47
  {
48
  "path": "config_sentence_transformers.json",
 
74
  "sha256": "84e40c8e006c9b1d6c122e02cba9b02458120b5fb0c87b746c41e0207cf642cf",
75
  "size_bytes": 349
76
  },
77
+ {
78
+ "path": "runtime_audit.json",
79
+ "sha256": "abf8214273c17b842205bad9015e23388a2b81aa8c44b08506c562e57f9babf0",
80
+ "size_bytes": 2016
81
+ },
82
  {
83
  "path": "tokenizer.json",
84
  "sha256": "93623af029cdc69b87f2864d3b2cc2424fdf16684f15e139b5b9d08ec34ced91",
config.json CHANGED
The diff for this file is too large to render. See raw diff
 
runtime_audit.json ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "applied_config_corrections": [
3
+ {
4
+ "field": "quantization.mode",
5
+ "new": "mxfp8",
6
+ "old": "affine",
7
+ "physical_evidence_consistent": true
8
+ },
9
+ {
10
+ "field": "quantization_config.mode",
11
+ "new": "mxfp8",
12
+ "old": "affine",
13
+ "physical_evidence_consistent": true
14
+ }
15
+ ],
16
+ "current_config_sha256": "b49450c5c373f46fa637822bbbd466ada245fb3b58c40e589d41a8932070670f",
17
+ "date": "2026-10-06",
18
+ "header_sha256": {
19
+ "model.safetensors": "d6b3f9419188abb3890994fbfd1f6982179b14b3a634951169520b52e46d2fdb"
20
+ },
21
+ "input_bindings": {
22
+ "axquant_manifest.json": "fbb3ac237972ada8e6f5d7d4f7f7c13aff9c346d18cd8b96a7ef6b6d45aa5944",
23
+ "config.json": "3aeb949fad423de8a2d7c0c6d9bb52baf18a1c92e72b5761040f31ff1a3eebd0",
24
+ "model.safetensors.index.json": "6ec8c2db4004f31b9d945957229ba7619a59734e07778e42e70c4b583880a88f"
25
+ },
26
+ "issues": [],
27
+ "model_type": "qwen3",
28
+ "mtp_files": [],
29
+ "ngram_action": "none; do not invent n-gram data",
30
+ "ngram_files": [],
31
+ "ngram_quantization": [],
32
+ "ngram_table_metadata": null,
33
+ "ngram_tensor_count": 0,
34
+ "quality_certified": false,
35
+ "quantization": {
36
+ "quantization": {
37
+ "container_mode": "affine",
38
+ "per_module_modes": {
39
+ "mxfp8": 197
40
+ },
41
+ "physical_recipe_issues": []
42
+ },
43
+ "quantization_config": {
44
+ "container_mode": "affine",
45
+ "per_module_modes": {
46
+ "mxfp8": 197
47
+ },
48
+ "physical_recipe_issues": []
49
+ }
50
+ },
51
+ "removed_source_evidence": [],
52
+ "repo_id": "AutomatosX/AX-Qwen3-Embedding-0.6B-MLX-AXQ-MXFP8",
53
+ "runtime_arch_id": null,
54
+ "runtime_verified": false,
55
+ "schema_version": "axquant.hub-runtime-audit.v1",
56
+ "scope": "Pinned remote config/index/Safetensors header audit; no runtime load or generation claim.",
57
+ "source_revision": "dadc5445b6069b0179c323f11ec93026942cb622",
58
+ "status": "no-audited-peer-export-profile",
59
+ "unchanged_weight_sha256": {
60
+ "model.safetensors": "4ff6f3b73633a5194cf0ba50b944789466eb47a4a2b199f5db8944a566fd22d8"
61
+ }
62
+ }