Hugging Face's logo

aws-neuron
/
optimum-neuron-cache

Model card Files
xet
 Community
658
dacorvo HF Staff commited on
Commit
69b3867
·
verified ·
1 Parent(s): 6cac8d3

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 +12 -0
  2. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev3/gemma3_text/unsloth/gemma-3-270m-it/e912866445bcdd7bd532.json +82 -0
  3. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev3/llama/unsloth/Llama-3.2-1B-Instruct/124119d9f41b205009a0.json +64 -0
  4. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev3/qwen2/Qwen/Qwen2.5-0.5B/21c3bc6d9846044edf76.json +83 -0
  5. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev3/smollm3/HuggingFaceTB/SmolLM3-3B/87f1aaa5ae1a31bb7adc.json +135 -0
  6. neuronxcc-2.21.33363.0+82129205/MODULE_06bd0f82c14784b0fd44+86b837c2/compile_flags.json +1 -0
  7. neuronxcc-2.21.33363.0+82129205/MODULE_06bd0f82c14784b0fd44+86b837c2/model.done +0 -0
  8. neuronxcc-2.21.33363.0+82129205/MODULE_06bd0f82c14784b0fd44+86b837c2/model.hlo_module.pb +3 -0
  9. neuronxcc-2.21.33363.0+82129205/MODULE_06bd0f82c14784b0fd44+86b837c2/model.neff +3 -0
  10. neuronxcc-2.21.33363.0+82129205/MODULE_06bd0f82c14784b0fd44+86b837c2/wrapped_neff.hlo +3 -0
  11. neuronxcc-2.21.33363.0+82129205/MODULE_10244305442015770634+e30acd3a/model.neff +0 -0
  12. neuronxcc-2.21.33363.0+82129205/MODULE_10606948783918825529+e30acd3a/model.neff +0 -0
  13. neuronxcc-2.21.33363.0+82129205/MODULE_10645643398657092095+e30acd3a/model.neff +0 -0
  14. neuronxcc-2.21.33363.0+82129205/MODULE_10661660426924300837+e30acd3a/model.neff +0 -0
  15. neuronxcc-2.21.33363.0+82129205/MODULE_10746122569655005679+e30acd3a/model.neff +0 -0
  16. neuronxcc-2.21.33363.0+82129205/MODULE_11086318750207148626+e30acd3a/model.neff +0 -0
  17. neuronxcc-2.21.33363.0+82129205/MODULE_11301370888592189606+e30acd3a/model.neff +1 -1
  18. neuronxcc-2.21.33363.0+82129205/MODULE_11909310600244571805+e30acd3a/model.neff +0 -0
  19. neuronxcc-2.21.33363.0+82129205/MODULE_11933818254123612383+e30acd3a/model.neff +0 -0
  20. neuronxcc-2.21.33363.0+82129205/MODULE_1210392327607194823+e30acd3a/model.neff +0 -0
  21. neuronxcc-2.21.33363.0+82129205/MODULE_12cc2229352c8e66e5ca+bcc5b4e4/compile_flags.json +1 -0
  22. neuronxcc-2.21.33363.0+82129205/MODULE_12cc2229352c8e66e5ca+bcc5b4e4/model.done +0 -0
  23. neuronxcc-2.21.33363.0+82129205/MODULE_12cc2229352c8e66e5ca+bcc5b4e4/model.hlo_module.pb +3 -0
  24. neuronxcc-2.21.33363.0+82129205/MODULE_12cc2229352c8e66e5ca+bcc5b4e4/model.neff +0 -0
  25. neuronxcc-2.21.33363.0+82129205/MODULE_13102010590082783346+e30acd3a/model.neff +0 -0
  26. neuronxcc-2.21.33363.0+82129205/MODULE_13475273426270778455+e30acd3a/model.neff +0 -0
  27. neuronxcc-2.21.33363.0+82129205/MODULE_13728813963059599796+e30acd3a/model.neff +0 -0
  28. neuronxcc-2.21.33363.0+82129205/MODULE_13783872956654054643+e30acd3a/model.neff +0 -0
  29. neuronxcc-2.21.33363.0+82129205/MODULE_14097776130612710282+e30acd3a/model.neff +0 -0
  30. neuronxcc-2.21.33363.0+82129205/MODULE_14155105943057125178+e30acd3a/model.neff +0 -0
  31. neuronxcc-2.21.33363.0+82129205/MODULE_14173682194645188821+e30acd3a/model.neff +0 -0
  32. neuronxcc-2.21.33363.0+82129205/MODULE_14274833313744358348+e30acd3a/model.neff +0 -0
  33. neuronxcc-2.21.33363.0+82129205/MODULE_14469716438804935215+e30acd3a/model.neff +0 -0
  34. neuronxcc-2.21.33363.0+82129205/MODULE_14657035910656097540+e30acd3a/model.neff +1 -1
  35. neuronxcc-2.21.33363.0+82129205/MODULE_14739672283771660808+e30acd3a/model.neff +0 -0
  36. neuronxcc-2.21.33363.0+82129205/MODULE_15128894797204950836+e30acd3a/model.neff +1 -1
  37. neuronxcc-2.21.33363.0+82129205/MODULE_15557971880365771457+e30acd3a/model.neff +0 -0
  38. neuronxcc-2.21.33363.0+82129205/MODULE_15564425602929126510+e30acd3a/model.neff +0 -0
  39. neuronxcc-2.21.33363.0+82129205/MODULE_15589838460896944293+e30acd3a/model.neff +0 -0
  40. neuronxcc-2.21.33363.0+82129205/MODULE_15654572849171857535+e30acd3a/model.neff +0 -0
  41. neuronxcc-2.21.33363.0+82129205/MODULE_15767383571209512795+e30acd3a/model.neff +0 -0
  42. neuronxcc-2.21.33363.0+82129205/MODULE_15888836045088309511+e30acd3a/model.neff +0 -0
  43. neuronxcc-2.21.33363.0+82129205/MODULE_15929451261464042997+e30acd3a/model.neff +0 -0
  44. neuronxcc-2.21.33363.0+82129205/MODULE_16016505958416521648+e30acd3a/model.neff +0 -0
  45. neuronxcc-2.21.33363.0+82129205/MODULE_16053163479112702088+e30acd3a/model.neff +0 -0
  46. neuronxcc-2.21.33363.0+82129205/MODULE_16204387852795926216+e30acd3a/model.neff +0 -0
  47. neuronxcc-2.21.33363.0+82129205/MODULE_1656719109221189948+e30acd3a/model.neff +0 -0
  48. neuronxcc-2.21.33363.0+82129205/MODULE_16599571375348449904+e30acd3a/model.neff +0 -0
  49. neuronxcc-2.21.33363.0+82129205/MODULE_1663757798483801648+e30acd3a/model.neff +0 -0
  50. neuronxcc-2.21.33363.0+82129205/MODULE_16738296820980389103+e30acd3a/model.neff +0 -0
.gitattributes CHANGED
@@ -17127,3 +17127,15 @@ neuronxcc-2.21.33363.0+82129205/MODULE_8eb90a77a3f704c8f98b+b5739779/model.neff
17127
  neuronxcc-2.21.33363.0+82129205/MODULE_8eb90a77a3f704c8f98b+b5739779/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
17128
  neuronxcc-2.21.33363.0+82129205/MODULE_959095166de1daed47d1+23644079/model.neff filter=lfs diff=lfs merge=lfs -text
17129
  neuronxcc-2.21.33363.0+82129205/MODULE_959095166de1daed47d1+23644079/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
 
 
 
 
 
 
 
17127
  neuronxcc-2.21.33363.0+82129205/MODULE_8eb90a77a3f704c8f98b+b5739779/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
17128
  neuronxcc-2.21.33363.0+82129205/MODULE_959095166de1daed47d1+23644079/model.neff filter=lfs diff=lfs merge=lfs -text
17129
  neuronxcc-2.21.33363.0+82129205/MODULE_959095166de1daed47d1+23644079/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
17130
+ neuronxcc-2.21.33363.0+82129205/MODULE_06bd0f82c14784b0fd44+86b837c2/model.neff filter=lfs diff=lfs merge=lfs -text
17131
+ neuronxcc-2.21.33363.0+82129205/MODULE_06bd0f82c14784b0fd44+86b837c2/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
17132
+ neuronxcc-2.21.33363.0+82129205/MODULE_470e323783e541da4829+4a882b02/model.neff filter=lfs diff=lfs merge=lfs -text
17133
+ neuronxcc-2.21.33363.0+82129205/MODULE_470e323783e541da4829+4a882b02/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
17134
+ neuronxcc-2.21.33363.0+82129205/MODULE_53e3b013e9149feef6c4+7b04ddc0/model.neff filter=lfs diff=lfs merge=lfs -text
17135
+ neuronxcc-2.21.33363.0+82129205/MODULE_53e3b013e9149feef6c4+7b04ddc0/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
17136
+ neuronxcc-2.21.33363.0+82129205/MODULE_69bf5337d3e3b7930fc1+e2c9caa6/model.neff filter=lfs diff=lfs merge=lfs -text
17137
+ neuronxcc-2.21.33363.0+82129205/MODULE_69bf5337d3e3b7930fc1+e2c9caa6/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
17138
+ neuronxcc-2.21.33363.0+82129205/MODULE_a4272eb02d670a6abbf6+f3fa1b4a/model.neff filter=lfs diff=lfs merge=lfs -text
17139
+ neuronxcc-2.21.33363.0+82129205/MODULE_a4272eb02d670a6abbf6+f3fa1b4a/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
17140
+ neuronxcc-2.21.33363.0+82129205/MODULE_a8f2e63c352bbdccaf42+6155346e/model.neff filter=lfs diff=lfs merge=lfs -text
17141
+ neuronxcc-2.21.33363.0+82129205/MODULE_a8f2e63c352bbdccaf42+6155346e/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev3/gemma3_text/unsloth/gemma-3-270m-it/e912866445bcdd7bd532.json ADDED
@@ -0,0 +1,82 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "unsloth/gemma-3-270m-it",
4
+ "_sliding_window_pattern": 6,
5
+ "_task": "text-generation",
6
+ "architectures": [
7
+ "Gemma3ForCausalLM"
8
+ ],
9
+ "attention_bias": false,
10
+ "attention_dropout": 0.0,
11
+ "attn_logit_softcapping": null,
12
+ "dtype": "bfloat16",
13
+ "final_logit_softcapping": null,
14
+ "head_dim": 256,
15
+ "hidden_activation": "gelu_pytorch_tanh",
16
+ "hidden_size": 640,
17
+ "initializer_range": 0.02,
18
+ "intermediate_size": 2048,
19
+ "layer_types": [
20
+ "sliding_attention",
21
+ "sliding_attention",
22
+ "sliding_attention",
23
+ "sliding_attention",
24
+ "sliding_attention",
25
+ "full_attention",
26
+ "sliding_attention",
27
+ "sliding_attention",
28
+ "sliding_attention",
29
+ "sliding_attention",
30
+ "sliding_attention",
31
+ "full_attention",
32
+ "sliding_attention",
33
+ "sliding_attention",
34
+ "sliding_attention",
35
+ "sliding_attention",
36
+ "sliding_attention",
37
+ "full_attention"
38
+ ],
39
+ "max_position_embeddings": 32768,
40
+ "model_type": "gemma3_text",
41
+ "neuron": {
42
+ "_serialized_key": "NxDNeuronConfig",
43
+ "batch_size": 1,
44
+ "capacity_factor": null,
45
+ "checkpoint_id": "unsloth/gemma-3-270m-it",
46
+ "checkpoint_revision": "23cf460f6bb16954176b3ddcc8d4f250501458a9",
47
+ "continuous_batching": false,
48
+ "ep_degree": 1,
49
+ "fused_qkv": true,
50
+ "glu_mlp": true,
51
+ "local_ranks_size": 2,
52
+ "max_batch_size": 1,
53
+ "max_context_length": 1024,
54
+ "max_topk": 256,
55
+ "n_active_tokens": 1024,
56
+ "neuronxcc_version": "2.21.33363.0+82129205",
57
+ "on_device_sampling": true,
58
+ "optimum_neuron_version": "0.4.6.dev3",
59
+ "output_logits": false,
60
+ "pp_degree": 1,
61
+ "prefill_chunk_size": 0,
62
+ "sequence_length": 1024,
63
+ "speculation_length": 0,
64
+ "start_rank_id": 0,
65
+ "target": "trn1",
66
+ "torch_dtype": "bfloat16",
67
+ "tp_degree": 2
68
+ },
69
+ "num_attention_heads": 4,
70
+ "num_hidden_layers": 18,
71
+ "num_key_value_heads": 1,
72
+ "query_pre_attn_scalar": 256,
73
+ "rms_norm_eps": 1e-06,
74
+ "rope_local_base_freq": 10000.0,
75
+ "rope_scaling": null,
76
+ "rope_theta": 1000000.0,
77
+ "sliding_window": 512,
78
+ "unsloth_fixed": true,
79
+ "use_bidirectional_attention": false,
80
+ "use_cache": true,
81
+ "vocab_size": 262144
82
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev3/llama/unsloth/Llama-3.2-1B-Instruct/124119d9f41b205009a0.json ADDED
@@ -0,0 +1,64 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "unsloth/Llama-3.2-1B-Instruct",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "dtype": "bfloat16",
11
+ "head_dim": 64,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 2048,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 8192,
16
+ "max_position_embeddings": 131072,
17
+ "mlp_bias": false,
18
+ "model_type": "llama",
19
+ "neuron": {
20
+ "_serialized_key": "NxDNeuronConfig",
21
+ "batch_size": 1,
22
+ "capacity_factor": null,
23
+ "checkpoint_id": "unsloth/Llama-3.2-1B-Instruct",
24
+ "checkpoint_revision": "5a8abab4a5d6f164389b1079fb721cfab8d7126c",
25
+ "continuous_batching": false,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 2,
30
+ "max_batch_size": 1,
31
+ "max_context_length": 4096,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 4096,
34
+ "neuronxcc_version": "2.21.33363.0+82129205",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.6.dev3",
37
+ "output_logits": false,
38
+ "pp_degree": 1,
39
+ "prefill_chunk_size": 1024,
40
+ "sequence_length": 4096,
41
+ "speculation_length": 0,
42
+ "start_rank_id": 0,
43
+ "target": "trn1",
44
+ "torch_dtype": "bfloat16",
45
+ "tp_degree": 2
46
+ },
47
+ "num_attention_heads": 32,
48
+ "num_hidden_layers": 16,
49
+ "num_key_value_heads": 8,
50
+ "pretraining_tp": 1,
51
+ "rms_norm_eps": 1e-05,
52
+ "rope_scaling": {
53
+ "factor": 32.0,
54
+ "high_freq_factor": 4.0,
55
+ "low_freq_factor": 1.0,
56
+ "original_max_position_embeddings": 8192,
57
+ "rope_type": "llama3"
58
+ },
59
+ "rope_theta": 500000.0,
60
+ "tie_word_embeddings": true,
61
+ "unsloth_fixed": true,
62
+ "use_cache": true,
63
+ "vocab_size": 128256
64
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev3/qwen2/Qwen/Qwen2.5-0.5B/21c3bc6d9846044edf76.json ADDED
@@ -0,0 +1,83 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": 2,
46
+ "capacity_factor": null,
47
+ "checkpoint_id": "Qwen/Qwen2.5-0.5B",
48
+ "checkpoint_revision": "060db6499f32faf8b98477b0a26969ef7d8b9987",
49
+ "continuous_batching": true,
50
+ "ep_degree": 1,
51
+ "fused_qkv": false,
52
+ "glu_mlp": true,
53
+ "local_ranks_size": 2,
54
+ "max_batch_size": 2,
55
+ "max_context_length": 1024,
56
+ "max_topk": 256,
57
+ "n_active_tokens": 1024,
58
+ "neuronxcc_version": "2.21.33363.0+82129205",
59
+ "on_device_sampling": false,
60
+ "optimum_neuron_version": "0.4.6.dev3",
61
+ "output_logits": false,
62
+ "pp_degree": 1,
63
+ "prefill_chunk_size": 0,
64
+ "sequence_length": 1024,
65
+ "speculation_length": 0,
66
+ "start_rank_id": 0,
67
+ "target": "trn1",
68
+ "torch_dtype": "bfloat16",
69
+ "tp_degree": 2
70
+ },
71
+ "num_attention_heads": 14,
72
+ "num_hidden_layers": 24,
73
+ "num_key_value_heads": 2,
74
+ "rms_norm_eps": 1e-06,
75
+ "rope_scaling": null,
76
+ "rope_theta": 1000000.0,
77
+ "sliding_window": null,
78
+ "tie_word_embeddings": true,
79
+ "use_cache": true,
80
+ "use_mrope": false,
81
+ "use_sliding_window": false,
82
+ "vocab_size": 151936
83
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev3/smollm3/HuggingFaceTB/SmolLM3-3B/87f1aaa5ae1a31bb7adc.json ADDED
@@ -0,0 +1,135 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "HuggingFaceTB/SmolLM3-3B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "SmolLM3ForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "dtype": "bfloat16",
11
+ "hidden_act": "silu",
12
+ "hidden_size": 2048,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 11008,
15
+ "layer_types": [
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
+ "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
+ ],
53
+ "max_position_embeddings": 65536,
54
+ "max_window_layers": 28,
55
+ "mlp_bias": false,
56
+ "model_type": "smollm3",
57
+ "neuron": {
58
+ "_serialized_key": "NxDNeuronConfig",
59
+ "batch_size": 1,
60
+ "capacity_factor": null,
61
+ "checkpoint_id": "HuggingFaceTB/SmolLM3-3B",
62
+ "checkpoint_revision": "a07cc9a04f16550a088caea529712d1d335b0ac1",
63
+ "continuous_batching": false,
64
+ "ep_degree": 1,
65
+ "fused_qkv": true,
66
+ "glu_mlp": true,
67
+ "local_ranks_size": 2,
68
+ "max_batch_size": 1,
69
+ "max_context_length": 1024,
70
+ "max_topk": 256,
71
+ "n_active_tokens": 1024,
72
+ "neuronxcc_version": "2.21.33363.0+82129205",
73
+ "on_device_sampling": true,
74
+ "optimum_neuron_version": "0.4.6.dev3",
75
+ "output_logits": false,
76
+ "pp_degree": 1,
77
+ "prefill_chunk_size": 0,
78
+ "sequence_length": 1024,
79
+ "speculation_length": 0,
80
+ "start_rank_id": 0,
81
+ "target": "trn1",
82
+ "torch_dtype": "bfloat16",
83
+ "tp_degree": 2
84
+ },
85
+ "no_rope_layer_interval": 4,
86
+ "no_rope_layers": [
87
+ 1,
88
+ 1,
89
+ 1,
90
+ 0,
91
+ 1,
92
+ 1,
93
+ 1,
94
+ 0,
95
+ 1,
96
+ 1,
97
+ 1,
98
+ 0,
99
+ 1,
100
+ 1,
101
+ 1,
102
+ 0,
103
+ 1,
104
+ 1,
105
+ 1,
106
+ 0,
107
+ 1,
108
+ 1,
109
+ 1,
110
+ 0,
111
+ 1,
112
+ 1,
113
+ 1,
114
+ 0,
115
+ 1,
116
+ 1,
117
+ 1,
118
+ 0,
119
+ 1,
120
+ 1,
121
+ 1,
122
+ 0
123
+ ],
124
+ "num_attention_heads": 16,
125
+ "num_hidden_layers": 36,
126
+ "num_key_value_heads": 4,
127
+ "pretraining_tp": 2,
128
+ "rms_norm_eps": 1e-06,
129
+ "rope_scaling": null,
130
+ "rope_theta": 5000000.0,
131
+ "sliding_window": null,
132
+ "use_cache": false,
133
+ "use_sliding_window": false,
134
+ "vocab_size": 128256
135
+ }
neuronxcc-2.21.33363.0+82129205/MODULE_06bd0f82c14784b0fd44+86b837c2/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--enable-saturate-infinity", "--auto-cast=none", "--model-type=transformer", "-O1", "--logfile=/tmp/nxdi_test_f708c9e7-1c92-4522-934e-2e3ded139c58/compiler_workdir/AttentionModelWrapper/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
neuronxcc-2.21.33363.0+82129205/MODULE_06bd0f82c14784b0fd44+86b837c2/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_06bd0f82c14784b0fd44+86b837c2/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0eb674f0ac78f8d22ded7c755c983de3edb92563308d60b734d19e694f489568
3
+ size 11280
neuronxcc-2.21.33363.0+82129205/MODULE_06bd0f82c14784b0fd44+86b837c2/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:afaee406217a526bdaad81c16f8d38780c2da2dc07f507c18feb97f574110fab
3
+ size 1444864
neuronxcc-2.21.33363.0+82129205/MODULE_06bd0f82c14784b0fd44+86b837c2/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a399b58e4144baa7d15294a9b89c1ccb945d3f2250a263ecebe278cb7651a544
3
+ size 1447845
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_11301370888592189606+e30acd3a/model.neff CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0c9ea2a1608587632f8e58f8c2619eec3f946ebd846e75a9c3e4c1b32a799a35
3
  size 246784
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6b9c5fa212f6e53feeb051307f450d5edc181964fa89b7a5a4f61dab31dab132
3
  size 246784
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_12cc2229352c8e66e5ca+bcc5b4e4/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--enable-saturate-infinity", "--auto-cast=none", "--model-type=transformer", "-O1", "--logfile=/tmp/nxdi_test_7436405a-6720-46bd-9d26-c9a43fcecb2f/compiler_workdir/Llama3RotaryEmbedding/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
neuronxcc-2.21.33363.0+82129205/MODULE_12cc2229352c8e66e5ca+bcc5b4e4/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_12cc2229352c8e66e5ca+bcc5b4e4/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f4a29d4cf27aec9af90e53324eff9ee69e033679482f4016a9dd3e5416c7c68a
3
+ size 1584
neuronxcc-2.21.33363.0+82129205/MODULE_12cc2229352c8e66e5ca+bcc5b4e4/model.neff ADDED
Binary file (42 kB). View file
 
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_14657035910656097540+e30acd3a/model.neff CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7adf01ee775d350dfa5975c0e6c03f7cb5d51c49021907af57999cb614c380d1
3
  size 123904
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:71177301b95958f030c970d75b4dbb7520ff0ad505a36bf447da09df891e634b
3
  size 123904
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_15128894797204950836+e30acd3a/model.neff CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3eed4b70b26e4f84102dcd8b84a44b5e053ad5e0c1ef9fc81cab40166ee4f02f
3
  size 113664
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:441bf26cdc1ca746b11097252007267a54e1252bf0fcf202f473f3893479d623
3
  size 113664
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