dacorvo HF Staff commited on
Commit
8368b77
·
verified ·
1 Parent(s): b53df1e

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 +18 -0
  2. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev1/granite/ibm-granite/granite-3.1-2b-instruct/a55342f00a2b27869b7c.json +58 -0
  3. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev1/phi3/microsoft/Phi-3.5-mini-instruct/b4d345e70123e586f5ad.json +164 -0
  4. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev1/qwen2/Qwen/Qwen2.5-0.5B/5d91fecb93d8e1d0e384.json +82 -0
  5. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev1/qwen3/Qwen/Qwen3-0.6B/0bef133446631b35e4a1.json +87 -0
  6. neuronxcc-2.21.33363.0+82129205/MODULE_04d083e7ddd421914edd+a02c3a36/model.neff +1 -1
  7. neuronxcc-2.21.33363.0+82129205/MODULE_04d083e7ddd421914edd+a02c3a36/wrapped_neff.hlo +1 -1
  8. neuronxcc-2.21.33363.0+82129205/MODULE_10217061096959125489+e30acd3a/model.neff +0 -0
  9. neuronxcc-2.21.33363.0+82129205/MODULE_10244305442015770634+e30acd3a/model.neff +0 -0
  10. neuronxcc-2.21.33363.0+82129205/MODULE_10606948783918825529+e30acd3a/model.neff +0 -0
  11. neuronxcc-2.21.33363.0+82129205/MODULE_10645643398657092095+e30acd3a/model.neff +0 -0
  12. neuronxcc-2.21.33363.0+82129205/MODULE_10661660426924300837+e30acd3a/model.neff +0 -0
  13. neuronxcc-2.21.33363.0+82129205/MODULE_10746122569655005679+e30acd3a/model.neff +0 -0
  14. neuronxcc-2.21.33363.0+82129205/MODULE_11086318750207148626+e30acd3a/model.neff +0 -0
  15. neuronxcc-2.21.33363.0+82129205/MODULE_11909310600244571805+e30acd3a/model.neff +0 -0
  16. neuronxcc-2.21.33363.0+82129205/MODULE_11933818254123612383+e30acd3a/model.neff +0 -0
  17. neuronxcc-2.21.33363.0+82129205/MODULE_1210392327607194823+e30acd3a/model.neff +0 -0
  18. neuronxcc-2.21.33363.0+82129205/MODULE_12487216553200321032+e30acd3a/model.neff +0 -0
  19. neuronxcc-2.21.33363.0+82129205/MODULE_13085549342645515693+e30acd3a/model.neff +0 -0
  20. neuronxcc-2.21.33363.0+82129205/MODULE_13102010590082783346+e30acd3a/model.neff +0 -0
  21. neuronxcc-2.21.33363.0+82129205/MODULE_13475273426270778455+e30acd3a/model.neff +0 -0
  22. neuronxcc-2.21.33363.0+82129205/MODULE_13728813963059599796+e30acd3a/model.neff +0 -0
  23. neuronxcc-2.21.33363.0+82129205/MODULE_13783872956654054643+e30acd3a/model.neff +0 -0
  24. neuronxcc-2.21.33363.0+82129205/MODULE_14097776130612710282+e30acd3a/model.neff +0 -0
  25. neuronxcc-2.21.33363.0+82129205/MODULE_14155105943057125178+e30acd3a/model.neff +0 -0
  26. neuronxcc-2.21.33363.0+82129205/MODULE_14173682194645188821+e30acd3a/model.neff +0 -0
  27. neuronxcc-2.21.33363.0+82129205/MODULE_14274833313744358348+e30acd3a/model.neff +0 -0
  28. neuronxcc-2.21.33363.0+82129205/MODULE_14469716438804935215+e30acd3a/model.neff +0 -0
  29. neuronxcc-2.21.33363.0+82129205/MODULE_14739672283771660808+e30acd3a/model.neff +0 -0
  30. neuronxcc-2.21.33363.0+82129205/MODULE_15442663025941492357+e30acd3a/model.neff +0 -0
  31. neuronxcc-2.21.33363.0+82129205/MODULE_15557971880365771457+e30acd3a/model.neff +0 -0
  32. neuronxcc-2.21.33363.0+82129205/MODULE_15564425602929126510+e30acd3a/model.neff +0 -0
  33. neuronxcc-2.21.33363.0+82129205/MODULE_15589838460896944293+e30acd3a/model.neff +0 -0
  34. neuronxcc-2.21.33363.0+82129205/MODULE_15654572849171857535+e30acd3a/model.neff +0 -0
  35. neuronxcc-2.21.33363.0+82129205/MODULE_15767383571209512795+e30acd3a/model.neff +0 -0
  36. neuronxcc-2.21.33363.0+82129205/MODULE_15888836045088309511+e30acd3a/model.neff +0 -0
  37. neuronxcc-2.21.33363.0+82129205/MODULE_15929451261464042997+e30acd3a/model.neff +0 -0
  38. neuronxcc-2.21.33363.0+82129205/MODULE_16016505958416521648+e30acd3a/model.neff +0 -0
  39. neuronxcc-2.21.33363.0+82129205/MODULE_16053163479112702088+e30acd3a/model.neff +0 -0
  40. neuronxcc-2.21.33363.0+82129205/MODULE_16204387852795926216+e30acd3a/model.neff +0 -0
  41. neuronxcc-2.21.33363.0+82129205/MODULE_1656719109221189948+e30acd3a/model.neff +0 -0
  42. neuronxcc-2.21.33363.0+82129205/MODULE_16599571375348449904+e30acd3a/model.neff +0 -0
  43. neuronxcc-2.21.33363.0+82129205/MODULE_1663757798483801648+e30acd3a/model.neff +0 -0
  44. neuronxcc-2.21.33363.0+82129205/MODULE_16738296820980389103+e30acd3a/model.neff +0 -0
  45. neuronxcc-2.21.33363.0+82129205/MODULE_1696825468766062114+e30acd3a/model.neff +0 -0
  46. neuronxcc-2.21.33363.0+82129205/MODULE_17476209562158013765+e30acd3a/model.neff +0 -0
  47. neuronxcc-2.21.33363.0+82129205/MODULE_17781596253725927902+e30acd3a/model.neff +0 -0
  48. neuronxcc-2.21.33363.0+82129205/MODULE_1795740353831178306+e30acd3a/model.neff +0 -0
  49. neuronxcc-2.21.33363.0+82129205/MODULE_1800832390737682969+e30acd3a/model.neff +0 -0
  50. neuronxcc-2.21.33363.0+82129205/MODULE_1805481651134498710+e30acd3a/model.neff +0 -0
.gitattributes CHANGED
@@ -14817,3 +14817,21 @@ neuronxcc-2.21.33363.0+82129205/MODULE_e9f45339826914e7685d+573f850f/wrapped_nef
14817
  neuronxcc-2.22.12471.0+b4a00d10/MODULE_73645ab8f4c76b2a2ba7+84337dd9/model.neff filter=lfs diff=lfs merge=lfs -text
14818
  neuronxcc-2.22.12471.0+b4a00d10/MODULE_73645ab8f4c76b2a2ba7+84337dd9/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
14819
  neuronxcc-2.22.12471.0+b4a00d10/MODULE_7a9140b490a5784e886b+519f203d/model.neff filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
14817
  neuronxcc-2.22.12471.0+b4a00d10/MODULE_73645ab8f4c76b2a2ba7+84337dd9/model.neff filter=lfs diff=lfs merge=lfs -text
14818
  neuronxcc-2.22.12471.0+b4a00d10/MODULE_73645ab8f4c76b2a2ba7+84337dd9/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
14819
  neuronxcc-2.22.12471.0+b4a00d10/MODULE_7a9140b490a5784e886b+519f203d/model.neff filter=lfs diff=lfs merge=lfs -text
14820
+ neuronxcc-2.21.33363.0+82129205/MODULE_238f936f6792c3472f34+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
14821
+ neuronxcc-2.21.33363.0+82129205/MODULE_4b953d1c85671e655ef7+0ea7ca8a/model.neff filter=lfs diff=lfs merge=lfs -text
14822
+ neuronxcc-2.21.33363.0+82129205/MODULE_4b953d1c85671e655ef7+0ea7ca8a/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
14823
+ neuronxcc-2.21.33363.0+82129205/MODULE_78c9e4beae8ae2b514a7+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
14824
+ neuronxcc-2.21.33363.0+82129205/MODULE_84bf2c7d60809359ba58+a9e994e0/model.neff filter=lfs diff=lfs merge=lfs -text
14825
+ neuronxcc-2.21.33363.0+82129205/MODULE_84bf2c7d60809359ba58+a9e994e0/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
14826
+ neuronxcc-2.21.33363.0+82129205/MODULE_92d87119e7896084888f+c8418bf2/model.neff filter=lfs diff=lfs merge=lfs -text
14827
+ neuronxcc-2.21.33363.0+82129205/MODULE_92d87119e7896084888f+c8418bf2/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
14828
+ neuronxcc-2.21.33363.0+82129205/MODULE_9fe6e286c1c48b3cf2a9+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
14829
+ neuronxcc-2.21.33363.0+82129205/MODULE_9fe6e286c1c48b3cf2a9+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
14830
+ neuronxcc-2.21.33363.0+82129205/MODULE_be2371af345021274e80+d366e126/model.neff filter=lfs diff=lfs merge=lfs -text
14831
+ neuronxcc-2.21.33363.0+82129205/MODULE_be2371af345021274e80+d366e126/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
14832
+ neuronxcc-2.21.33363.0+82129205/MODULE_d21e7dca1cad4cb73da0+3a51130f/model.neff filter=lfs diff=lfs merge=lfs -text
14833
+ neuronxcc-2.21.33363.0+82129205/MODULE_d21e7dca1cad4cb73da0+3a51130f/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
14834
+ neuronxcc-2.21.33363.0+82129205/MODULE_f2ffb8f9fa7b301cf0e2+9c446665/model.neff filter=lfs diff=lfs merge=lfs -text
14835
+ neuronxcc-2.21.33363.0+82129205/MODULE_f2ffb8f9fa7b301cf0e2+9c446665/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
14836
+ neuronxcc-2.21.33363.0+82129205/MODULE_f6a3ff2ad752e338116c+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
14837
+ neuronxcc-2.21.33363.0+82129205/MODULE_f6a3ff2ad752e338116c+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev1/granite/ibm-granite/granite-3.1-2b-instruct/a55342f00a2b27869b7c.json ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "ibm-granite/granite-3.1-2b-instruct",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "GraniteForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.1,
10
+ "attention_multiplier": 0.015625,
11
+ "dtype": "bfloat16",
12
+ "embedding_multiplier": 12.0,
13
+ "hidden_act": "silu",
14
+ "hidden_size": 2048,
15
+ "initializer_range": 0.02,
16
+ "intermediate_size": 8192,
17
+ "logits_scaling": 8.0,
18
+ "max_position_embeddings": 131072,
19
+ "mlp_bias": false,
20
+ "model_type": "granite",
21
+ "neuron": {
22
+ "_serialized_key": "NxDNeuronConfig",
23
+ "batch_size": 1,
24
+ "capacity_factor": null,
25
+ "checkpoint_id": "ibm-granite/granite-3.1-2b-instruct",
26
+ "checkpoint_revision": "bbc2aed595bd38bd770263dc3ab831db9794441d",
27
+ "continuous_batching": false,
28
+ "ep_degree": 1,
29
+ "fused_qkv": true,
30
+ "glu_mlp": true,
31
+ "local_ranks_size": 2,
32
+ "max_batch_size": 1,
33
+ "max_context_length": 8192,
34
+ "max_topk": 256,
35
+ "n_active_tokens": 8192,
36
+ "neuronxcc_version": "2.21.33363.0+82129205",
37
+ "on_device_sampling": true,
38
+ "optimum_neuron_version": "0.4.4.dev1",
39
+ "output_logits": false,
40
+ "pp_degree": 1,
41
+ "sequence_length": 8192,
42
+ "speculation_length": 0,
43
+ "start_rank_id": 0,
44
+ "target": "trn1",
45
+ "torch_dtype": "bfloat16",
46
+ "tp_degree": 2
47
+ },
48
+ "num_attention_heads": 32,
49
+ "num_hidden_layers": 40,
50
+ "num_key_value_heads": 8,
51
+ "residual_multiplier": 0.22,
52
+ "rms_norm_eps": 1e-05,
53
+ "rope_scaling": null,
54
+ "rope_theta": 5000000.0,
55
+ "tie_word_embeddings": true,
56
+ "use_cache": true,
57
+ "vocab_size": 49155
58
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev1/phi3/microsoft/Phi-3.5-mini-instruct/b4d345e70123e586f5ad.json ADDED
@@ -0,0 +1,164 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "microsoft/Phi-3.5-mini-instruct",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Phi3ForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "auto_map": {
11
+ "AutoConfig": "configuration_phi3.Phi3Config",
12
+ "AutoModelForCausalLM": "modeling_phi3.Phi3ForCausalLM"
13
+ },
14
+ "dtype": "bfloat16",
15
+ "embd_pdrop": 0.0,
16
+ "hidden_act": "silu",
17
+ "hidden_size": 3072,
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 8192,
20
+ "max_position_embeddings": 131072,
21
+ "model_type": "phi3",
22
+ "neuron": {
23
+ "_serialized_key": "NxDNeuronConfig",
24
+ "batch_size": 4,
25
+ "capacity_factor": null,
26
+ "checkpoint_id": "microsoft/Phi-3.5-mini-instruct",
27
+ "checkpoint_revision": "2fe192450127e6a83f7441aef6e3ca586c338b77",
28
+ "continuous_batching": true,
29
+ "ep_degree": 1,
30
+ "fused_qkv": true,
31
+ "glu_mlp": true,
32
+ "local_ranks_size": 2,
33
+ "max_batch_size": 4,
34
+ "max_context_length": 4096,
35
+ "max_topk": 256,
36
+ "n_active_tokens": 4096,
37
+ "neuronxcc_version": "2.21.33363.0+82129205",
38
+ "on_device_sampling": true,
39
+ "optimum_neuron_version": "0.4.4.dev1",
40
+ "output_logits": false,
41
+ "pp_degree": 1,
42
+ "sequence_length": 4096,
43
+ "speculation_length": 0,
44
+ "start_rank_id": 0,
45
+ "target": "trn1",
46
+ "torch_dtype": "bfloat16",
47
+ "tp_degree": 2
48
+ },
49
+ "num_attention_heads": 32,
50
+ "num_hidden_layers": 32,
51
+ "num_key_value_heads": 32,
52
+ "original_max_position_embeddings": 4096,
53
+ "partial_rotary_factor": 1.0,
54
+ "resid_pdrop": 0.0,
55
+ "rms_norm_eps": 1e-05,
56
+ "rope_scaling": {
57
+ "long_factor": [
58
+ 1.0800000429153442,
59
+ 1.1100000143051147,
60
+ 1.1399999856948853,
61
+ 1.340000033378601,
62
+ 1.5899999141693115,
63
+ 1.600000023841858,
64
+ 1.6200000047683716,
65
+ 2.620000123977661,
66
+ 3.2300000190734863,
67
+ 3.2300000190734863,
68
+ 4.789999961853027,
69
+ 7.400000095367432,
70
+ 7.700000286102295,
71
+ 9.09000015258789,
72
+ 12.199999809265137,
73
+ 17.670000076293945,
74
+ 24.46000099182129,
75
+ 28.57000160217285,
76
+ 30.420001983642578,
77
+ 30.840002059936523,
78
+ 32.590003967285156,
79
+ 32.93000411987305,
80
+ 42.320003509521484,
81
+ 44.96000289916992,
82
+ 50.340003967285156,
83
+ 50.45000457763672,
84
+ 57.55000305175781,
85
+ 57.93000411987305,
86
+ 58.21000289916992,
87
+ 60.1400032043457,
88
+ 62.61000442504883,
89
+ 62.62000274658203,
90
+ 62.71000289916992,
91
+ 63.1400032043457,
92
+ 63.1400032043457,
93
+ 63.77000427246094,
94
+ 63.93000411987305,
95
+ 63.96000289916992,
96
+ 63.970001220703125,
97
+ 64.02999877929688,
98
+ 64.06999969482422,
99
+ 64.08000183105469,
100
+ 64.12000274658203,
101
+ 64.41000366210938,
102
+ 64.4800033569336,
103
+ 64.51000213623047,
104
+ 64.52999877929688,
105
+ 64.83999633789062
106
+ ],
107
+ "short_factor": [
108
+ 1.0,
109
+ 1.0199999809265137,
110
+ 1.0299999713897705,
111
+ 1.0299999713897705,
112
+ 1.0499999523162842,
113
+ 1.0499999523162842,
114
+ 1.0499999523162842,
115
+ 1.0499999523162842,
116
+ 1.0499999523162842,
117
+ 1.0699999332427979,
118
+ 1.0999999046325684,
119
+ 1.1099998950958252,
120
+ 1.1599998474121094,
121
+ 1.1599998474121094,
122
+ 1.1699998378753662,
123
+ 1.2899998426437378,
124
+ 1.339999794960022,
125
+ 1.679999828338623,
126
+ 1.7899998426437378,
127
+ 1.8199998140335083,
128
+ 1.8499997854232788,
129
+ 1.8799997568130493,
130
+ 1.9099997282028198,
131
+ 1.9399996995925903,
132
+ 1.9899996519088745,
133
+ 2.0199997425079346,
134
+ 2.0199997425079346,
135
+ 2.0199997425079346,
136
+ 2.0199997425079346,
137
+ 2.0199997425079346,
138
+ 2.0199997425079346,
139
+ 2.0299997329711914,
140
+ 2.0299997329711914,
141
+ 2.0299997329711914,
142
+ 2.0299997329711914,
143
+ 2.0299997329711914,
144
+ 2.0299997329711914,
145
+ 2.0299997329711914,
146
+ 2.0299997329711914,
147
+ 2.0299997329711914,
148
+ 2.0799996852874756,
149
+ 2.0899996757507324,
150
+ 2.189999580383301,
151
+ 2.2199995517730713,
152
+ 2.5899994373321533,
153
+ 2.729999542236328,
154
+ 2.749999523162842,
155
+ 2.8399994373321533
156
+ ],
157
+ "type": "longrope"
158
+ },
159
+ "rope_theta": 10000.0,
160
+ "sliding_window": 262144,
161
+ "tie_word_embeddings": false,
162
+ "use_cache": true,
163
+ "vocab_size": 32064
164
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev1/qwen2/Qwen/Qwen2.5-0.5B/5d91fecb93d8e1d0e384.json ADDED
@@ -0,0 +1,82 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "Qwen/Qwen2.5-0.5B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Qwen2ForCausalLM"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "dtype": "bfloat16",
10
+ "hidden_act": "silu",
11
+ "hidden_size": 896,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 4864,
14
+ "layer_types": [
15
+ "full_attention",
16
+ "full_attention",
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
+ ],
40
+ "max_position_embeddings": 32768,
41
+ "max_window_layers": 24,
42
+ "model_type": "qwen2",
43
+ "neuron": {
44
+ "_serialized_key": "NxDNeuronConfig",
45
+ "batch_size": 1,
46
+ "capacity_factor": null,
47
+ "checkpoint_id": "Qwen/Qwen2.5-0.5B",
48
+ "checkpoint_revision": "060db6499f32faf8b98477b0a26969ef7d8b9987",
49
+ "continuous_batching": false,
50
+ "ep_degree": 1,
51
+ "fused_qkv": false,
52
+ "glu_mlp": true,
53
+ "local_ranks_size": 2,
54
+ "max_batch_size": 1,
55
+ "max_context_length": 8192,
56
+ "max_topk": 256,
57
+ "n_active_tokens": 8192,
58
+ "neuronxcc_version": "2.21.33363.0+82129205",
59
+ "on_device_sampling": true,
60
+ "optimum_neuron_version": "0.4.4.dev1",
61
+ "output_logits": false,
62
+ "pp_degree": 1,
63
+ "sequence_length": 8192,
64
+ "speculation_length": 0,
65
+ "start_rank_id": 0,
66
+ "target": "trn1",
67
+ "torch_dtype": "bfloat16",
68
+ "tp_degree": 2
69
+ },
70
+ "num_attention_heads": 14,
71
+ "num_hidden_layers": 24,
72
+ "num_key_value_heads": 2,
73
+ "rms_norm_eps": 1e-06,
74
+ "rope_scaling": null,
75
+ "rope_theta": 1000000.0,
76
+ "sliding_window": null,
77
+ "tie_word_embeddings": true,
78
+ "use_cache": true,
79
+ "use_mrope": false,
80
+ "use_sliding_window": false,
81
+ "vocab_size": 151936
82
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev1/qwen3/Qwen/Qwen3-0.6B/0bef133446631b35e4a1.json ADDED
@@ -0,0 +1,87 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "Qwen/Qwen3-0.6B",
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": 40960,
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/Qwen3-0.6B",
54
+ "checkpoint_revision": "c1899de289a04d12100db370d81485cdf75e47ca",
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": "0.4.4.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": 151936
87
+ }
neuronxcc-2.21.33363.0+82129205/MODULE_04d083e7ddd421914edd+a02c3a36/model.neff CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b2f2e1fd51f5a6cc33ca4f98bf96f931433e6bf7fcc58f11194f6cd583681432
3
  size 2100224
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9dbb3a703b7487958838b3469e4e94811bac90646ced5c0096e55d5fab40ae65
3
  size 2100224
neuronxcc-2.21.33363.0+82129205/MODULE_04d083e7ddd421914edd+a02c3a36/wrapped_neff.hlo CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c1bf1adbd4dd98e2a5bc158dd7a3844a3ea3e7fd33b9182c1bd50ae1af57c6cc
3
  size 2174311
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:67ba0facf639555663df7923b2e69d01bded1590e1e62b5d2a77a47072a3d261
3
  size 2174311
neuronxcc-2.21.33363.0+82129205/MODULE_10217061096959125489+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_10217061096959125489+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_10217061096959125489+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_10244305442015770634+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_10244305442015770634+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_10244305442015770634+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_10606948783918825529+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_10606948783918825529+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_10606948783918825529+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_10645643398657092095+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_10645643398657092095+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_10645643398657092095+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_10661660426924300837+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_10661660426924300837+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_10661660426924300837+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_10746122569655005679+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_10746122569655005679+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_10746122569655005679+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_11086318750207148626+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_11086318750207148626+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_11086318750207148626+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_11909310600244571805+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_11909310600244571805+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_11909310600244571805+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_11933818254123612383+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_11933818254123612383+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_11933818254123612383+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_1210392327607194823+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_1210392327607194823+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_1210392327607194823+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_12487216553200321032+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_12487216553200321032+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_12487216553200321032+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_13085549342645515693+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_13085549342645515693+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_13085549342645515693+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_13102010590082783346+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_13102010590082783346+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_13102010590082783346+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_13475273426270778455+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_13475273426270778455+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_13475273426270778455+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_13728813963059599796+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_13728813963059599796+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_13728813963059599796+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_13783872956654054643+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_13783872956654054643+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_13783872956654054643+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_14097776130612710282+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_14097776130612710282+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_14097776130612710282+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_14155105943057125178+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_14155105943057125178+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_14155105943057125178+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_14173682194645188821+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_14173682194645188821+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_14173682194645188821+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_14274833313744358348+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_14274833313744358348+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_14274833313744358348+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_14469716438804935215+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_14469716438804935215+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_14469716438804935215+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_14739672283771660808+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_14739672283771660808+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_14739672283771660808+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_15442663025941492357+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_15442663025941492357+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_15442663025941492357+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_15557971880365771457+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_15557971880365771457+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_15557971880365771457+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_15564425602929126510+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_15564425602929126510+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_15564425602929126510+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_15589838460896944293+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_15589838460896944293+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_15589838460896944293+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_15654572849171857535+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_15654572849171857535+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_15654572849171857535+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_15767383571209512795+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_15767383571209512795+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_15767383571209512795+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_15888836045088309511+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_15888836045088309511+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_15888836045088309511+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_15929451261464042997+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_15929451261464042997+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_15929451261464042997+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_16016505958416521648+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_16016505958416521648+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_16016505958416521648+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_16053163479112702088+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_16053163479112702088+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_16053163479112702088+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_16204387852795926216+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_16204387852795926216+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_16204387852795926216+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_1656719109221189948+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_1656719109221189948+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_1656719109221189948+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_16599571375348449904+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_16599571375348449904+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_16599571375348449904+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_1663757798483801648+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_1663757798483801648+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_1663757798483801648+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_16738296820980389103+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_16738296820980389103+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_16738296820980389103+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_1696825468766062114+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_1696825468766062114+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_1696825468766062114+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_17476209562158013765+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_17476209562158013765+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_17476209562158013765+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_17781596253725927902+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_17781596253725927902+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_17781596253725927902+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_1795740353831178306+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_1795740353831178306+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_1795740353831178306+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_1800832390737682969+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_1800832390737682969+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_1800832390737682969+e30acd3a/model.neff differ
 
neuronxcc-2.21.33363.0+82129205/MODULE_1805481651134498710+e30acd3a/model.neff CHANGED
Binary files a/neuronxcc-2.21.33363.0+82129205/MODULE_1805481651134498710+e30acd3a/model.neff and b/neuronxcc-2.21.33363.0+82129205/MODULE_1805481651134498710+e30acd3a/model.neff differ