Hugging Face's logo

aws-neuron
/
optimum-neuron-cache

Model card Files
xet
Community
647
dacorvo HF Staff commited on
Commit
7b9dd22
·
verified ·
1 Parent(s): 66983f0

Synchronizing local compiler cache.

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. .gitattributes +25 -0
  2. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev1/4d281af1651143f9a83b42af25d9f0a533dbbd9c3441537127ae772fd3e399cc/0dfe723550263376ebbf.json +87 -0
  3. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev1/4d281af1651143f9a83b42af25d9f0a533dbbd9c3441537127ae772fd3e399cc/468de8d6877b1b5efb87.json +87 -0
  4. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev1/4d281af1651143f9a83b42af25d9f0a533dbbd9c3441537127ae772fd3e399cc/c5697348fd90121b22ca.json +87 -0
  5. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev1/4d281af1651143f9a83b42af25d9f0a533dbbd9c3441537127ae772fd3e399cc/dbf0d3cf63a4b9c18134.json +87 -0
  6. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev1/4d281af1651143f9a83b42af25d9f0a533dbbd9c3441537127ae772fd3e399cc/e14467c71bee5cf6b4df.json +87 -0
  7. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev1/qwen3/Qwen/Qwen3-Embedding-0.6B/dbf0d3cf63a4b9c18134.json +87 -0
  8. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev2/60feecaa0c4c075e2f3e46a3f55d9a273f0ddd75a0ecf64e4ae27352e0819506/1c4ee5d7dc71b8843fca.json +87 -0
  9. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev2/e0b6d1e2424243dcd9ff1755e02969dcc312d14df531d876c5c2892f285b2863/07552dc6c695df3ea557.json +95 -0
  10. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev2/qwen3/Qwen/Qwen3-Embedding-0.6B/1c4ee5d7dc71b8843fca.json +87 -0
  11. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev2/qwen3/Qwen/Qwen3-Embedding-4B/07552dc6c695df3ea557.json +95 -0
  12. neuronxcc-2.21.33363.0+82129205/MODULE_003d3d9afcec48e81ba4+a02c3a36/model.neff +1 -1
  13. neuronxcc-2.21.33363.0+82129205/MODULE_003d3d9afcec48e81ba4+a02c3a36/wrapped_neff.hlo +1 -1
  14. neuronxcc-2.21.33363.0+82129205/MODULE_01328268a19023ed0f24+f8e6d902/model.hlo_module.pb +1 -1
  15. neuronxcc-2.21.33363.0+82129205/MODULE_01328268a19023ed0f24+f8e6d902/model.neff +1 -1
  16. neuronxcc-2.21.33363.0+82129205/MODULE_065f10c75a3a003852db+fb4cc044/compile_flags.json +1 -0
  17. neuronxcc-2.21.33363.0+82129205/MODULE_065f10c75a3a003852db+fb4cc044/model.done +0 -0
  18. neuronxcc-2.21.33363.0+82129205/MODULE_065f10c75a3a003852db+fb4cc044/model.hlo_module.pb +3 -0
  19. neuronxcc-2.21.33363.0+82129205/MODULE_065f10c75a3a003852db+fb4cc044/model.neff +3 -0
  20. neuronxcc-2.21.33363.0+82129205/MODULE_0a5cbcb0f0d605182b40+24129607/compile_flags.json +1 -0
  21. neuronxcc-2.21.33363.0+82129205/MODULE_0a5cbcb0f0d605182b40+24129607/model.hlo_module.pb +3 -0
  22. neuronxcc-2.21.33363.0+82129205/MODULE_0a5cbcb0f0d605182b40+24129607/model.log +1 -0
  23. neuronxcc-2.21.33363.0+82129205/MODULE_0e8f2cac7949cfdcbe7c+a02c3a36/compile_flags.json +1 -0
  24. neuronxcc-2.21.33363.0+82129205/MODULE_0e8f2cac7949cfdcbe7c+a02c3a36/model.done +0 -0
  25. neuronxcc-2.21.33363.0+82129205/MODULE_0e8f2cac7949cfdcbe7c+a02c3a36/model.hlo_module.pb +3 -0
  26. neuronxcc-2.21.33363.0+82129205/MODULE_0e8f2cac7949cfdcbe7c+a02c3a36/model.neff +3 -0
  27. neuronxcc-2.21.33363.0+82129205/MODULE_0e8f2cac7949cfdcbe7c+a02c3a36/wrapped_neff.hlo +3 -0
  28. neuronxcc-2.21.33363.0+82129205/MODULE_1934c2fc93e7793f0aef+a02c3a36/model.hlo_module.pb +1 -1
  29. neuronxcc-2.21.33363.0+82129205/MODULE_1934c2fc93e7793f0aef+a02c3a36/model.neff +1 -1
  30. neuronxcc-2.21.33363.0+82129205/MODULE_1934c2fc93e7793f0aef+a02c3a36/wrapped_neff.hlo +1 -1
  31. neuronxcc-2.21.33363.0+82129205/MODULE_1944f672afeaa6480c61+24129607/model.hlo_module.pb +1 -1
  32. neuronxcc-2.21.33363.0+82129205/MODULE_1944f672afeaa6480c61+24129607/model.neff +1 -1
  33. neuronxcc-2.21.33363.0+82129205/MODULE_1aace3454acab4662925+fb4cc044/compile_flags.json +1 -0
  34. neuronxcc-2.21.33363.0+82129205/MODULE_1aace3454acab4662925+fb4cc044/model.done +0 -0
  35. neuronxcc-2.21.33363.0+82129205/MODULE_1aace3454acab4662925+fb4cc044/model.hlo_module.pb +3 -0
  36. neuronxcc-2.21.33363.0+82129205/MODULE_1aace3454acab4662925+fb4cc044/model.neff +3 -0
  37. neuronxcc-2.21.33363.0+82129205/MODULE_1cc0ec46877f3aa47ae6+24129607/compile_flags.json +1 -0
  38. neuronxcc-2.21.33363.0+82129205/MODULE_1cc0ec46877f3aa47ae6+24129607/model.done +0 -0
  39. neuronxcc-2.21.33363.0+82129205/MODULE_1cc0ec46877f3aa47ae6+24129607/model.hlo_module.pb +3 -0
  40. neuronxcc-2.21.33363.0+82129205/MODULE_1cc0ec46877f3aa47ae6+24129607/model.neff +3 -0
  41. neuronxcc-2.21.33363.0+82129205/MODULE_21fc076779532e6e1e41+24129607/compile_flags.json +1 -0
  42. neuronxcc-2.21.33363.0+82129205/MODULE_21fc076779532e6e1e41+24129607/model.done +0 -0
  43. neuronxcc-2.21.33363.0+82129205/MODULE_21fc076779532e6e1e41+24129607/model.hlo_module.pb +3 -0
  44. neuronxcc-2.21.33363.0+82129205/MODULE_21fc076779532e6e1e41+24129607/model.neff +3 -0
  45. neuronxcc-2.21.33363.0+82129205/MODULE_250ecf58b3a2e7ac5757+24129607/model.hlo_module.pb +1 -1
  46. neuronxcc-2.21.33363.0+82129205/MODULE_250ecf58b3a2e7ac5757+24129607/model.neff +1 -1
  47. neuronxcc-2.21.33363.0+82129205/MODULE_2672b897bf6327bea9e4+24129607/model.hlo_module.pb +1 -1
  48. neuronxcc-2.21.33363.0+82129205/MODULE_2672b897bf6327bea9e4+24129607/model.neff +1 -1
  49. neuronxcc-2.21.33363.0+82129205/MODULE_27dcb29f0a28b1d6f51e+a02c3a36/compile_flags.json +1 -0
  50. neuronxcc-2.21.33363.0+82129205/MODULE_27dcb29f0a28b1d6f51e+a02c3a36/model.done +0 -0
.gitattributes CHANGED
@@ -16108,3 +16108,28 @@ neuronxcc-2.21.33363.0+82129205/MODULE_efea8e54de072c4916ed+a32116a7/model.neff
16108
  neuronxcc-2.21.33363.0+82129205/MODULE_119a20f83dc85c879ef7+ac10809c/model.neff filter=lfs diff=lfs merge=lfs -text
16109
  neuronxcc-2.21.33363.0+82129205/MODULE_119a20f83dc85c879ef7+ac10809c/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
16110
  neuronxcc-2.21.33363.0+82129205/MODULE_1d1ee05517b3b870704a+a32116a7/model.neff filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
16108
  neuronxcc-2.21.33363.0+82129205/MODULE_119a20f83dc85c879ef7+ac10809c/model.neff filter=lfs diff=lfs merge=lfs -text
16109
  neuronxcc-2.21.33363.0+82129205/MODULE_119a20f83dc85c879ef7+ac10809c/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
16110
  neuronxcc-2.21.33363.0+82129205/MODULE_1d1ee05517b3b870704a+a32116a7/model.neff filter=lfs diff=lfs merge=lfs -text
16111
+ neuronxcc-2.21.33363.0+82129205/MODULE_065f10c75a3a003852db+fb4cc044/model.neff filter=lfs diff=lfs merge=lfs -text
16112
+ neuronxcc-2.21.33363.0+82129205/MODULE_0e8f2cac7949cfdcbe7c+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
16113
+ neuronxcc-2.21.33363.0+82129205/MODULE_0e8f2cac7949cfdcbe7c+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
16114
+ neuronxcc-2.21.33363.0+82129205/MODULE_1aace3454acab4662925+fb4cc044/model.neff filter=lfs diff=lfs merge=lfs -text
16115
+ neuronxcc-2.21.33363.0+82129205/MODULE_1cc0ec46877f3aa47ae6+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
16116
+ neuronxcc-2.21.33363.0+82129205/MODULE_21fc076779532e6e1e41+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
16117
+ neuronxcc-2.21.33363.0+82129205/MODULE_27dcb29f0a28b1d6f51e+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
16118
+ neuronxcc-2.21.33363.0+82129205/MODULE_27dcb29f0a28b1d6f51e+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
16119
+ neuronxcc-2.21.33363.0+82129205/MODULE_2828c4ae6bc360cc555f+fb4cc044/model.neff filter=lfs diff=lfs merge=lfs -text
16120
+ neuronxcc-2.21.33363.0+82129205/MODULE_4cf1e5f6e4c17bb56bb0+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
16121
+ neuronxcc-2.21.33363.0+82129205/MODULE_4cf1e5f6e4c17bb56bb0+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
16122
+ neuronxcc-2.21.33363.0+82129205/MODULE_55c2fa803a7e3881cef6+fb4cc044/model.neff filter=lfs diff=lfs merge=lfs -text
16123
+ neuronxcc-2.21.33363.0+82129205/MODULE_79eb3abf001c934aa416+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
16124
+ neuronxcc-2.21.33363.0+82129205/MODULE_9b4ea40b364ed3edd3ad+fb4cc044/model.neff filter=lfs diff=lfs merge=lfs -text
16125
+ neuronxcc-2.21.33363.0+82129205/MODULE_a17ecef0a02d21e4d5a8+fb4cc044/model.neff filter=lfs diff=lfs merge=lfs -text
16126
+ neuronxcc-2.21.33363.0+82129205/MODULE_a8c9bb79a74a03571fcd+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
16127
+ neuronxcc-2.21.33363.0+82129205/MODULE_b1b042cf3d88bb04c2e0+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
16128
+ neuronxcc-2.21.33363.0+82129205/MODULE_b1b042cf3d88bb04c2e0+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
16129
+ neuronxcc-2.21.33363.0+82129205/MODULE_b46fffa5d4b2e69904a7+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
16130
+ neuronxcc-2.21.33363.0+82129205/MODULE_b46fffa5d4b2e69904a7+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
16131
+ neuronxcc-2.21.33363.0+82129205/MODULE_b8bea2e3eb787b70e571+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
16132
+ neuronxcc-2.21.33363.0+82129205/MODULE_b8bea2e3eb787b70e571+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
16133
+ neuronxcc-2.21.33363.0+82129205/MODULE_cb86ed7fc724b06726b3+fb4cc044/model.neff filter=lfs diff=lfs merge=lfs -text
16134
+ neuronxcc-2.21.33363.0+82129205/MODULE_d3f277cf573c91b9ded8+fb4cc044/model.neff filter=lfs diff=lfs merge=lfs -text
16135
+ neuronxcc-2.21.33363.0+82129205/MODULE_dd01d7a2913cbd8c1ba1+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev1/4d281af1651143f9a83b42af25d9f0a533dbbd9c3441537127ae772fd3e399cc/0dfe723550263376ebbf.json ADDED
@@ -0,0 +1,87 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "Qwen/Qwen3-Embedding-0.6B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Qwen3ForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "dtype": "bfloat16",
11
+ "head_dim": 128,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 1024,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 3072,
16
+ "layer_types": [
17
+ "full_attention",
18
+ "full_attention",
19
+ "full_attention",
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention"
45
+ ],
46
+ "max_position_embeddings": 32768,
47
+ "max_window_layers": 28,
48
+ "model_type": "qwen3",
49
+ "neuron": {
50
+ "_serialized_key": "NxDNeuronConfig",
51
+ "batch_size": 4,
52
+ "capacity_factor": null,
53
+ "checkpoint_id": "Qwen/Qwen3-Embedding-0.6B",
54
+ "checkpoint_revision": "c54f2e6e80b2d7b7de06f51cec4959f6b3e03418",
55
+ "continuous_batching": true,
56
+ "ep_degree": 1,
57
+ "fused_qkv": true,
58
+ "glu_mlp": true,
59
+ "local_ranks_size": 2,
60
+ "max_batch_size": 4,
61
+ "max_context_length": 4096,
62
+ "max_topk": 256,
63
+ "n_active_tokens": 4096,
64
+ "neuronxcc_version": "2.21.33363.0+82129205",
65
+ "on_device_sampling": false,
66
+ "optimum_neuron_version": "0.4.5.dev1",
67
+ "output_logits": false,
68
+ "pp_degree": 1,
69
+ "sequence_length": 4096,
70
+ "speculation_length": 0,
71
+ "start_rank_id": 0,
72
+ "target": "trn1",
73
+ "torch_dtype": "bfloat16",
74
+ "tp_degree": 2
75
+ },
76
+ "num_attention_heads": 16,
77
+ "num_hidden_layers": 28,
78
+ "num_key_value_heads": 8,
79
+ "rms_norm_eps": 1e-06,
80
+ "rope_scaling": null,
81
+ "rope_theta": 1000000,
82
+ "sliding_window": null,
83
+ "tie_word_embeddings": true,
84
+ "use_cache": true,
85
+ "use_sliding_window": false,
86
+ "vocab_size": 151669
87
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev1/4d281af1651143f9a83b42af25d9f0a533dbbd9c3441537127ae772fd3e399cc/468de8d6877b1b5efb87.json ADDED
@@ -0,0 +1,87 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "Qwen/Qwen3-Embedding-0.6B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Qwen3ForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "dtype": "bfloat16",
11
+ "head_dim": 128,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 1024,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 3072,
16
+ "layer_types": [
17
+ "full_attention",
18
+ "full_attention",
19
+ "full_attention",
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention"
45
+ ],
46
+ "max_position_embeddings": 32768,
47
+ "max_window_layers": 28,
48
+ "model_type": "qwen3",
49
+ "neuron": {
50
+ "_serialized_key": "NxDNeuronConfig",
51
+ "batch_size": 1,
52
+ "capacity_factor": null,
53
+ "checkpoint_id": "Qwen/Qwen3-Embedding-0.6B",
54
+ "checkpoint_revision": "c54f2e6e80b2d7b7de06f51cec4959f6b3e03418",
55
+ "continuous_batching": false,
56
+ "ep_degree": 1,
57
+ "fused_qkv": true,
58
+ "glu_mlp": true,
59
+ "local_ranks_size": 2,
60
+ "max_batch_size": 1,
61
+ "max_context_length": 8192,
62
+ "max_topk": 256,
63
+ "n_active_tokens": 8192,
64
+ "neuronxcc_version": "2.21.33363.0+82129205",
65
+ "on_device_sampling": true,
66
+ "optimum_neuron_version": "0.4.5.dev1",
67
+ "output_logits": false,
68
+ "pp_degree": 1,
69
+ "sequence_length": 8192,
70
+ "speculation_length": 0,
71
+ "start_rank_id": 0,
72
+ "target": "trn1",
73
+ "torch_dtype": "bfloat16",
74
+ "tp_degree": 2
75
+ },
76
+ "num_attention_heads": 16,
77
+ "num_hidden_layers": 28,
78
+ "num_key_value_heads": 8,
79
+ "rms_norm_eps": 1e-06,
80
+ "rope_scaling": null,
81
+ "rope_theta": 1000000,
82
+ "sliding_window": null,
83
+ "tie_word_embeddings": true,
84
+ "use_cache": true,
85
+ "use_sliding_window": false,
86
+ "vocab_size": 151669
87
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev1/4d281af1651143f9a83b42af25d9f0a533dbbd9c3441537127ae772fd3e399cc/c5697348fd90121b22ca.json ADDED
@@ -0,0 +1,87 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "Qwen/Qwen3-Embedding-0.6B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Qwen3ForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "dtype": "bfloat16",
11
+ "head_dim": 128,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 1024,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 3072,
16
+ "layer_types": [
17
+ "full_attention",
18
+ "full_attention",
19
+ "full_attention",
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention"
45
+ ],
46
+ "max_position_embeddings": 32768,
47
+ "max_window_layers": 28,
48
+ "model_type": "qwen3",
49
+ "neuron": {
50
+ "_serialized_key": "NxDNeuronConfig",
51
+ "batch_size": 6,
52
+ "capacity_factor": null,
53
+ "checkpoint_id": "Qwen/Qwen3-Embedding-0.6B",
54
+ "checkpoint_revision": "c54f2e6e80b2d7b7de06f51cec4959f6b3e03418",
55
+ "continuous_batching": true,
56
+ "ep_degree": 1,
57
+ "fused_qkv": true,
58
+ "glu_mlp": true,
59
+ "local_ranks_size": 2,
60
+ "max_batch_size": 6,
61
+ "max_context_length": 8192,
62
+ "max_topk": 256,
63
+ "n_active_tokens": 8192,
64
+ "neuronxcc_version": "2.21.33363.0+82129205",
65
+ "on_device_sampling": false,
66
+ "optimum_neuron_version": "0.4.5.dev1",
67
+ "output_logits": false,
68
+ "pp_degree": 1,
69
+ "sequence_length": 8192,
70
+ "speculation_length": 0,
71
+ "start_rank_id": 0,
72
+ "target": "trn1",
73
+ "torch_dtype": "bfloat16",
74
+ "tp_degree": 2
75
+ },
76
+ "num_attention_heads": 16,
77
+ "num_hidden_layers": 28,
78
+ "num_key_value_heads": 8,
79
+ "rms_norm_eps": 1e-06,
80
+ "rope_scaling": null,
81
+ "rope_theta": 1000000,
82
+ "sliding_window": null,
83
+ "tie_word_embeddings": true,
84
+ "use_cache": true,
85
+ "use_sliding_window": false,
86
+ "vocab_size": 151669
87
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev1/4d281af1651143f9a83b42af25d9f0a533dbbd9c3441537127ae772fd3e399cc/dbf0d3cf63a4b9c18134.json ADDED
@@ -0,0 +1,87 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "Qwen/Qwen3-Embedding-0.6B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Qwen3ForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "dtype": "bfloat16",
11
+ "head_dim": 128,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 1024,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 3072,
16
+ "layer_types": [
17
+ "full_attention",
18
+ "full_attention",
19
+ "full_attention",
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention"
45
+ ],
46
+ "max_position_embeddings": 32768,
47
+ "max_window_layers": 28,
48
+ "model_type": "qwen3",
49
+ "neuron": {
50
+ "_serialized_key": "NxDNeuronConfig",
51
+ "batch_size": 6,
52
+ "capacity_factor": null,
53
+ "checkpoint_id": "Qwen/Qwen3-Embedding-0.6B",
54
+ "checkpoint_revision": "c54f2e6e80b2d7b7de06f51cec4959f6b3e03418",
55
+ "continuous_batching": true,
56
+ "ep_degree": 1,
57
+ "fused_qkv": true,
58
+ "glu_mlp": true,
59
+ "local_ranks_size": 2,
60
+ "max_batch_size": 6,
61
+ "max_context_length": 1024,
62
+ "max_topk": 256,
63
+ "n_active_tokens": 1024,
64
+ "neuronxcc_version": "2.21.33363.0+82129205",
65
+ "on_device_sampling": false,
66
+ "optimum_neuron_version": "0.4.5.dev1",
67
+ "output_logits": false,
68
+ "pp_degree": 1,
69
+ "sequence_length": 1024,
70
+ "speculation_length": 0,
71
+ "start_rank_id": 0,
72
+ "target": "trn1",
73
+ "torch_dtype": "bfloat16",
74
+ "tp_degree": 2
75
+ },
76
+ "num_attention_heads": 16,
77
+ "num_hidden_layers": 28,
78
+ "num_key_value_heads": 8,
79
+ "rms_norm_eps": 1e-06,
80
+ "rope_scaling": null,
81
+ "rope_theta": 1000000,
82
+ "sliding_window": null,
83
+ "tie_word_embeddings": true,
84
+ "use_cache": true,
85
+ "use_sliding_window": false,
86
+ "vocab_size": 151669
87
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev1/4d281af1651143f9a83b42af25d9f0a533dbbd9c3441537127ae772fd3e399cc/e14467c71bee5cf6b4df.json ADDED
@@ -0,0 +1,87 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "Qwen/Qwen3-Embedding-0.6B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Qwen3ForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "dtype": "bfloat16",
11
+ "head_dim": 128,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 1024,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 3072,
16
+ "layer_types": [
17
+ "full_attention",
18
+ "full_attention",
19
+ "full_attention",
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention"
45
+ ],
46
+ "max_position_embeddings": 32768,
47
+ "max_window_layers": 28,
48
+ "model_type": "qwen3",
49
+ "neuron": {
50
+ "_serialized_key": "NxDNeuronConfig",
51
+ "batch_size": 4,
52
+ "capacity_factor": null,
53
+ "checkpoint_id": "Qwen/Qwen3-Embedding-0.6B",
54
+ "checkpoint_revision": "c54f2e6e80b2d7b7de06f51cec4959f6b3e03418",
55
+ "continuous_batching": true,
56
+ "ep_degree": 1,
57
+ "fused_qkv": true,
58
+ "glu_mlp": true,
59
+ "local_ranks_size": 2,
60
+ "max_batch_size": 4,
61
+ "max_context_length": 8192,
62
+ "max_topk": 256,
63
+ "n_active_tokens": 8192,
64
+ "neuronxcc_version": "2.21.33363.0+82129205",
65
+ "on_device_sampling": false,
66
+ "optimum_neuron_version": "0.4.5.dev1",
67
+ "output_logits": false,
68
+ "pp_degree": 1,
69
+ "sequence_length": 8192,
70
+ "speculation_length": 0,
71
+ "start_rank_id": 0,
72
+ "target": "trn1",
73
+ "torch_dtype": "bfloat16",
74
+ "tp_degree": 2
75
+ },
76
+ "num_attention_heads": 16,
77
+ "num_hidden_layers": 28,
78
+ "num_key_value_heads": 8,
79
+ "rms_norm_eps": 1e-06,
80
+ "rope_scaling": null,
81
+ "rope_theta": 1000000,
82
+ "sliding_window": null,
83
+ "tie_word_embeddings": true,
84
+ "use_cache": true,
85
+ "use_sliding_window": false,
86
+ "vocab_size": 151669
87
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev1/qwen3/Qwen/Qwen3-Embedding-0.6B/dbf0d3cf63a4b9c18134.json ADDED
@@ -0,0 +1,87 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "Qwen/Qwen3-Embedding-0.6B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Qwen3ForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "dtype": "bfloat16",
11
+ "head_dim": 128,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 1024,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 3072,
16
+ "layer_types": [
17
+ "full_attention",
18
+ "full_attention",
19
+ "full_attention",
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention"
45
+ ],
46
+ "max_position_embeddings": 32768,
47
+ "max_window_layers": 28,
48
+ "model_type": "qwen3",
49
+ "neuron": {
50
+ "_serialized_key": "NxDNeuronConfig",
51
+ "batch_size": 6,
52
+ "capacity_factor": null,
53
+ "checkpoint_id": "Qwen/Qwen3-Embedding-0.6B",
54
+ "checkpoint_revision": "c54f2e6e80b2d7b7de06f51cec4959f6b3e03418",
55
+ "continuous_batching": true,
56
+ "ep_degree": 1,
57
+ "fused_qkv": true,
58
+ "glu_mlp": true,
59
+ "local_ranks_size": 2,
60
+ "max_batch_size": 6,
61
+ "max_context_length": 1024,
62
+ "max_topk": 256,
63
+ "n_active_tokens": 1024,
64
+ "neuronxcc_version": "2.21.33363.0+82129205",
65
+ "on_device_sampling": false,
66
+ "optimum_neuron_version": "0.4.5.dev1",
67
+ "output_logits": false,
68
+ "pp_degree": 1,
69
+ "sequence_length": 1024,
70
+ "speculation_length": 0,
71
+ "start_rank_id": 0,
72
+ "target": "trn1",
73
+ "torch_dtype": "bfloat16",
74
+ "tp_degree": 2
75
+ },
76
+ "num_attention_heads": 16,
77
+ "num_hidden_layers": 28,
78
+ "num_key_value_heads": 8,
79
+ "rms_norm_eps": 1e-06,
80
+ "rope_scaling": null,
81
+ "rope_theta": 1000000,
82
+ "sliding_window": null,
83
+ "tie_word_embeddings": true,
84
+ "use_cache": true,
85
+ "use_sliding_window": false,
86
+ "vocab_size": 151669
87
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev2/60feecaa0c4c075e2f3e46a3f55d9a273f0ddd75a0ecf64e4ae27352e0819506/1c4ee5d7dc71b8843fca.json ADDED
@@ -0,0 +1,87 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "Qwen/Qwen3-Embedding-0.6B",
4
+ "_task": "feature-extraction",
5
+ "architectures": [
6
+ "Qwen3ForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "dtype": "bfloat16",
11
+ "head_dim": 128,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 1024,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 3072,
16
+ "layer_types": [
17
+ "full_attention",
18
+ "full_attention",
19
+ "full_attention",
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention"
45
+ ],
46
+ "max_position_embeddings": 32768,
47
+ "max_window_layers": 28,
48
+ "model_type": "qwen3",
49
+ "neuron": {
50
+ "_serialized_key": "NxDNeuronConfig",
51
+ "batch_size": 1,
52
+ "capacity_factor": null,
53
+ "checkpoint_id": "Qwen/Qwen3-Embedding-0.6B",
54
+ "checkpoint_revision": "c54f2e6e80b2d7b7de06f51cec4959f6b3e03418",
55
+ "continuous_batching": false,
56
+ "ep_degree": 1,
57
+ "fused_qkv": true,
58
+ "glu_mlp": true,
59
+ "local_ranks_size": 1,
60
+ "max_batch_size": 1,
61
+ "max_context_length": 8192,
62
+ "max_topk": 256,
63
+ "n_active_tokens": 8192,
64
+ "neuronxcc_version": "2.21.33363.0+82129205",
65
+ "on_device_sampling": false,
66
+ "optimum_neuron_version": "0.4.5.dev2",
67
+ "output_logits": false,
68
+ "pp_degree": 1,
69
+ "sequence_length": 8192,
70
+ "speculation_length": 0,
71
+ "start_rank_id": 0,
72
+ "target": "trn1",
73
+ "torch_dtype": "bfloat16",
74
+ "tp_degree": 1
75
+ },
76
+ "num_attention_heads": 16,
77
+ "num_hidden_layers": 28,
78
+ "num_key_value_heads": 8,
79
+ "rms_norm_eps": 1e-06,
80
+ "rope_scaling": null,
81
+ "rope_theta": 1000000,
82
+ "sliding_window": null,
83
+ "tie_word_embeddings": true,
84
+ "use_cache": true,
85
+ "use_sliding_window": false,
86
+ "vocab_size": 151669
87
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev2/e0b6d1e2424243dcd9ff1755e02969dcc312d14df531d876c5c2892f285b2863/07552dc6c695df3ea557.json ADDED
@@ -0,0 +1,95 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "Qwen/Qwen3-Embedding-4B",
4
+ "_task": "feature-extraction",
5
+ "architectures": [
6
+ "Qwen3ForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "dtype": "bfloat16",
11
+ "head_dim": 128,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 2560,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 9728,
16
+ "layer_types": [
17
+ "full_attention",
18
+ "full_attention",
19
+ "full_attention",
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention",
45
+ "full_attention",
46
+ "full_attention",
47
+ "full_attention",
48
+ "full_attention",
49
+ "full_attention",
50
+ "full_attention",
51
+ "full_attention",
52
+ "full_attention"
53
+ ],
54
+ "max_position_embeddings": 40960,
55
+ "max_window_layers": 36,
56
+ "model_type": "qwen3",
57
+ "neuron": {
58
+ "_serialized_key": "NxDNeuronConfig",
59
+ "batch_size": 1,
60
+ "capacity_factor": null,
61
+ "checkpoint_id": "Qwen/Qwen3-Embedding-4B",
62
+ "checkpoint_revision": "5cf2132abc99cad020ac570b19d031efec650f2b",
63
+ "continuous_batching": false,
64
+ "ep_degree": 1,
65
+ "fused_qkv": true,
66
+ "glu_mlp": true,
67
+ "local_ranks_size": 1,
68
+ "max_batch_size": 1,
69
+ "max_context_length": 4096,
70
+ "max_topk": 256,
71
+ "n_active_tokens": 4096,
72
+ "neuronxcc_version": "2.21.33363.0+82129205",
73
+ "on_device_sampling": false,
74
+ "optimum_neuron_version": "0.4.5.dev2",
75
+ "output_logits": false,
76
+ "pp_degree": 1,
77
+ "sequence_length": 4096,
78
+ "speculation_length": 0,
79
+ "start_rank_id": 0,
80
+ "target": "trn1",
81
+ "torch_dtype": "bfloat16",
82
+ "tp_degree": 1
83
+ },
84
+ "num_attention_heads": 32,
85
+ "num_hidden_layers": 36,
86
+ "num_key_value_heads": 8,
87
+ "rms_norm_eps": 1e-06,
88
+ "rope_scaling": null,
89
+ "rope_theta": 1000000,
90
+ "sliding_window": null,
91
+ "tie_word_embeddings": true,
92
+ "use_cache": true,
93
+ "use_sliding_window": false,
94
+ "vocab_size": 151665
95
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev2/qwen3/Qwen/Qwen3-Embedding-0.6B/1c4ee5d7dc71b8843fca.json ADDED
@@ -0,0 +1,87 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "Qwen/Qwen3-Embedding-0.6B",
4
+ "_task": "feature-extraction",
5
+ "architectures": [
6
+ "Qwen3ForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "dtype": "bfloat16",
11
+ "head_dim": 128,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 1024,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 3072,
16
+ "layer_types": [
17
+ "full_attention",
18
+ "full_attention",
19
+ "full_attention",
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention"
45
+ ],
46
+ "max_position_embeddings": 32768,
47
+ "max_window_layers": 28,
48
+ "model_type": "qwen3",
49
+ "neuron": {
50
+ "_serialized_key": "NxDNeuronConfig",
51
+ "batch_size": 1,
52
+ "capacity_factor": null,
53
+ "checkpoint_id": "Qwen/Qwen3-Embedding-0.6B",
54
+ "checkpoint_revision": "c54f2e6e80b2d7b7de06f51cec4959f6b3e03418",
55
+ "continuous_batching": false,
56
+ "ep_degree": 1,
57
+ "fused_qkv": true,
58
+ "glu_mlp": true,
59
+ "local_ranks_size": 1,
60
+ "max_batch_size": 1,
61
+ "max_context_length": 8192,
62
+ "max_topk": 256,
63
+ "n_active_tokens": 8192,
64
+ "neuronxcc_version": "2.21.33363.0+82129205",
65
+ "on_device_sampling": false,
66
+ "optimum_neuron_version": "0.4.5.dev2",
67
+ "output_logits": false,
68
+ "pp_degree": 1,
69
+ "sequence_length": 8192,
70
+ "speculation_length": 0,
71
+ "start_rank_id": 0,
72
+ "target": "trn1",
73
+ "torch_dtype": "bfloat16",
74
+ "tp_degree": 1
75
+ },
76
+ "num_attention_heads": 16,
77
+ "num_hidden_layers": 28,
78
+ "num_key_value_heads": 8,
79
+ "rms_norm_eps": 1e-06,
80
+ "rope_scaling": null,
81
+ "rope_theta": 1000000,
82
+ "sliding_window": null,
83
+ "tie_word_embeddings": true,
84
+ "use_cache": true,
85
+ "use_sliding_window": false,
86
+ "vocab_size": 151669
87
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev2/qwen3/Qwen/Qwen3-Embedding-4B/07552dc6c695df3ea557.json ADDED
@@ -0,0 +1,95 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "Qwen/Qwen3-Embedding-4B",
4
+ "_task": "feature-extraction",
5
+ "architectures": [
6
+ "Qwen3ForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "dtype": "bfloat16",
11
+ "head_dim": 128,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 2560,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 9728,
16
+ "layer_types": [
17
+ "full_attention",
18
+ "full_attention",
19
+ "full_attention",
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention",
45
+ "full_attention",
46
+ "full_attention",
47
+ "full_attention",
48
+ "full_attention",
49
+ "full_attention",
50
+ "full_attention",
51
+ "full_attention",
52
+ "full_attention"
53
+ ],
54
+ "max_position_embeddings": 40960,
55
+ "max_window_layers": 36,
56
+ "model_type": "qwen3",
57
+ "neuron": {
58
+ "_serialized_key": "NxDNeuronConfig",
59
+ "batch_size": 1,
60
+ "capacity_factor": null,
61
+ "checkpoint_id": "Qwen/Qwen3-Embedding-4B",
62
+ "checkpoint_revision": "5cf2132abc99cad020ac570b19d031efec650f2b",
63
+ "continuous_batching": false,
64
+ "ep_degree": 1,
65
+ "fused_qkv": true,
66
+ "glu_mlp": true,
67
+ "local_ranks_size": 1,
68
+ "max_batch_size": 1,
69
+ "max_context_length": 4096,
70
+ "max_topk": 256,
71
+ "n_active_tokens": 4096,
72
+ "neuronxcc_version": "2.21.33363.0+82129205",
73
+ "on_device_sampling": false,
74
+ "optimum_neuron_version": "0.4.5.dev2",
75
+ "output_logits": false,
76
+ "pp_degree": 1,
77
+ "sequence_length": 4096,
78
+ "speculation_length": 0,
79
+ "start_rank_id": 0,
80
+ "target": "trn1",
81
+ "torch_dtype": "bfloat16",
82
+ "tp_degree": 1
83
+ },
84
+ "num_attention_heads": 32,
85
+ "num_hidden_layers": 36,
86
+ "num_key_value_heads": 8,
87
+ "rms_norm_eps": 1e-06,
88
+ "rope_scaling": null,
89
+ "rope_theta": 1000000,
90
+ "sliding_window": null,
91
+ "tie_word_embeddings": true,
92
+ "use_cache": true,
93
+ "use_sliding_window": false,
94
+ "vocab_size": 151665
95
+ }
neuronxcc-2.21.33363.0+82129205/MODULE_003d3d9afcec48e81ba4+a02c3a36/model.neff CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:9773e8a89bdb9d5fee0d87e79df3820ba15a88789076f3b4252aebfc5bc43deb
3
  size 3503104
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dc81f037061f94e0f1fe3bc5bfffd4b5fdf4af0c261ea6c5614688d745efd2f4
3
  size 3503104
neuronxcc-2.21.33363.0+82129205/MODULE_003d3d9afcec48e81ba4+a02c3a36/wrapped_neff.hlo CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:535c1e201a5c612381965a264ba0ef64089144dd65205ec97f664971d65b498f
3
  size 3640161
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d92da39d5f156be691d57f5d60e9338aa9e80b18cc4d267ddd41a8d32e1768cc
3
  size 3640161
neuronxcc-2.21.33363.0+82129205/MODULE_01328268a19023ed0f24+f8e6d902/model.hlo_module.pb CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:5a7be97d2012aa41b5f56bf5bf86357a0dae77fd27821fa3e0737aba2fe5521e
3
  size 97794
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8ef50e479badb0312e208336533316393fc797857eb5a2cde744753c266d7f3f
3
  size 97794
neuronxcc-2.21.33363.0+82129205/MODULE_01328268a19023ed0f24+f8e6d902/model.neff CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7ff6c1f2a8f76328f3ffd5cd9cbb7ad62552ac54af96be0b68b754f3e79f2c6e
3
  size 410624
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:671f2968ee09229cb96bb22be152df8bd197829d6ce7ad77ef854a3963b716b9
3
  size 410624
neuronxcc-2.21.33363.0+82129205/MODULE_065f10c75a3a003852db+fb4cc044/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=1", "--logfile=/tmp/nxd_model/encoding/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.21.33363.0+82129205/MODULE_065f10c75a3a003852db+fb4cc044/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_065f10c75a3a003852db+fb4cc044/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1a60932659847098e53497491131db9516488a7e3c1c8ecc184ba9bbe4e3fb9a
3
+ size 662166
neuronxcc-2.21.33363.0+82129205/MODULE_065f10c75a3a003852db+fb4cc044/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4aa75cf62fd51441ac2a86aac870181fe1e178d38f055c50431a749169deef59
3
+ size 18484224
neuronxcc-2.21.33363.0+82129205/MODULE_0a5cbcb0f0d605182b40+24129607/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=1", "--logfile=/tmp/nxd_model/context_encoding/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.21.33363.0+82129205/MODULE_0a5cbcb0f0d605182b40+24129607/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7b6460ecb77b0df76fab544cee5f5c477229fd70e65cd0377841933f5251ae36
3
+ size 1134023
neuronxcc-2.21.33363.0+82129205/MODULE_0a5cbcb0f0d605182b40+24129607/model.log ADDED
@@ -0,0 +1 @@
 
 
1
+ Failed compilation with ['neuronx-cc', 'compile', '--framework=XLA', '/tmp/nxd_model/context_encoding/_tp0_bk0/model.MODULE_0a5cbcb0f0d605182b40+24129607.hlo_module.pb', '--output', '/tmp/nxd_model/context_encoding/_tp0_bk0/model.MODULE_0a5cbcb0f0d605182b40+24129607.neff', '--target=trn1', '--auto-cast=none', '--model-type=transformer', '--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ', '-O2', '--lnc=1', '--logfile=/tmp/nxd_model/context_encoding/_tp0_bk0/log-neuron-cc.txt', '--verbose=35']:
neuronxcc-2.21.33363.0+82129205/MODULE_0e8f2cac7949cfdcbe7c+a02c3a36/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=1", "--logfile=/tmp/nxd_model/token_generation/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
neuronxcc-2.21.33363.0+82129205/MODULE_0e8f2cac7949cfdcbe7c+a02c3a36/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_0e8f2cac7949cfdcbe7c+a02c3a36/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2480c10b0a667a77bc9f969b614c1c1627c7c0b72a47251ae70598baa2b8ef8c
3
+ size 1160172
neuronxcc-2.21.33363.0+82129205/MODULE_0e8f2cac7949cfdcbe7c+a02c3a36/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2dd6a56a6d6d101f9dc7e21594bc27ca343f95faa1e79f666426a8bc49de066f
3
+ size 5274624
neuronxcc-2.21.33363.0+82129205/MODULE_0e8f2cac7949cfdcbe7c+a02c3a36/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1965465ca9aae5d3723f6c6e5eb6986498f071af8bc72caec566f9e9dab8a8e8
3
+ size 5411681
neuronxcc-2.21.33363.0+82129205/MODULE_1934c2fc93e7793f0aef+a02c3a36/model.hlo_module.pb CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3ca0a9c8719d71e823180edfc6eba4a8842e74134383b6713cb51cd6134d5fb7
3
  size 596952
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c6361b23631f22b3534873d4bc3fde1c4197b34dc2d5d84b8097108cc7faed45
3
  size 596952
neuronxcc-2.21.33363.0+82129205/MODULE_1934c2fc93e7793f0aef+a02c3a36/model.neff CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:81f3a436a6014e92728a2af8e77d0d85eabeb5f40ccf61c11b4f0dcac0806751
3
  size 1936384
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0be47707e53a60759826fc077df760c768f27f3699574a1ca42a2c1721ac66bc
3
  size 1936384
neuronxcc-2.21.33363.0+82129205/MODULE_1934c2fc93e7793f0aef+a02c3a36/wrapped_neff.hlo CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c3ec97e51a42515c9ce13acbfba6965d41f7e0554d97ca9f9ab94caebd61aaa5
3
  size 2092575
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e23e21a3d39349b0a3b066fd29eb841c6bb9dba1fde90c7a0792c5717638a18
3
  size 2092575
neuronxcc-2.21.33363.0+82129205/MODULE_1944f672afeaa6480c61+24129607/model.hlo_module.pb CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:21b1204a4001c8a692290e9265478a1dbf162d41df0bba52b580fd22ab3a8735
3
  size 865460
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b77f0fd54de2b3f0e4d80ee388867e6ab1f396c5fd6c14be411f63d198649118
3
  size 865460
neuronxcc-2.21.33363.0+82129205/MODULE_1944f672afeaa6480c61+24129607/model.neff CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8e245ced68ece76c2e986e8d3bd5daf8aa2cb73ddc6d8aa87dfdf32196593303
3
  size 36148224
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bc4ee4b8b64d4dc6330fdd8fae2947adf4f88ba10b2505b8b51907e13863538d
3
  size 36148224
neuronxcc-2.21.33363.0+82129205/MODULE_1aace3454acab4662925+fb4cc044/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=1", "--logfile=/tmp/nxd_model/encoding/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.21.33363.0+82129205/MODULE_1aace3454acab4662925+fb4cc044/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_1aace3454acab4662925+fb4cc044/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ccd8dd07c208eb87030d40b6fe53666dbd98b0ad6fa3cf5fa46bd1e5d023563f
3
+ size 662166
neuronxcc-2.21.33363.0+82129205/MODULE_1aace3454acab4662925+fb4cc044/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:62694a86179db903be5f78ff966af1ec798da0d36a389aafbb60c4ae28079012
3
+ size 27689984
neuronxcc-2.21.33363.0+82129205/MODULE_1cc0ec46877f3aa47ae6+24129607/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=1", "--logfile=/tmp/nxd_model/context_encoding/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.21.33363.0+82129205/MODULE_1cc0ec46877f3aa47ae6+24129607/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_1cc0ec46877f3aa47ae6+24129607/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c14b363d2172922ab0e3ce4c160951e7e0305a0d219a88349452f0fb6def3a75
3
+ size 1305951
neuronxcc-2.21.33363.0+82129205/MODULE_1cc0ec46877f3aa47ae6+24129607/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:995e8864ea6a0be8f610b9cb61e158a28ed3a39eeb96f2160c3f43f0287d7dc4
3
+ size 98520064
neuronxcc-2.21.33363.0+82129205/MODULE_21fc076779532e6e1e41+24129607/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=1", "--logfile=/tmp/nxd_model/context_encoding/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.21.33363.0+82129205/MODULE_21fc076779532e6e1e41+24129607/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_21fc076779532e6e1e41+24129607/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0cfb81e51d1d924b03b36057b58cd756ac41d7388c73a9178ad22baccef42412
3
+ size 1305951
neuronxcc-2.21.33363.0+82129205/MODULE_21fc076779532e6e1e41+24129607/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9f7c100d4815aaffc226b9a25ef68cd5df5d7d3c333d97bfc35720818ad6bceb
3
+ size 36250624
neuronxcc-2.21.33363.0+82129205/MODULE_250ecf58b3a2e7ac5757+24129607/model.hlo_module.pb CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:329e621507127750dc610f95907a7e901219e2f34aaef4c17d140feef4a6c579
3
  size 1061093
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c1c161937d0eeeb2eca0dfcca2535ee497d38d1e83b3dbc45f21ccf5a0100888
3
  size 1061093
neuronxcc-2.21.33363.0+82129205/MODULE_250ecf58b3a2e7ac5757+24129607/model.neff CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:91ba54a6131873512c8412ed7a9972adca35b2f18f36e528de0b048bc3c788de
3
  size 9473024
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:af2c4a236a1f97d56e0840ca8099bf268c1fad8b15295cae8f906ae0c5534a0b
3
  size 9473024
neuronxcc-2.21.33363.0+82129205/MODULE_2672b897bf6327bea9e4+24129607/model.hlo_module.pb CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b4b6486a12378501e2484cf81b9f6885eabca76a7d817e3e2041c1736bac16d7
3
  size 82772
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6b56b0d331077542e72e5da19741dfcd8d5d6bdd01d35c1cdbe74227327956a6
3
  size 82772
neuronxcc-2.21.33363.0+82129205/MODULE_2672b897bf6327bea9e4+24129607/model.neff CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1c65b6b3a5f7c34ae231694981cf76347fb89393c25990695cd4f921f3d243e3
3
  size 267264
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d1cd6ca36926fad065cd480b6e705f247170aeb533426465d808212c1b7a5123
3
  size 267264
neuronxcc-2.21.33363.0+82129205/MODULE_27dcb29f0a28b1d6f51e+a02c3a36/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=1", "--logfile=/tmp/nxd_model/token_generation/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
neuronxcc-2.21.33363.0+82129205/MODULE_27dcb29f0a28b1d6f51e+a02c3a36/model.done ADDED
File without changes