Hugging Face's logo

optimum-internal-testing
/
neuron-testing-cache

Model card Files
xet
 Community
tengomucho HF Staff commited on
Commit
fa1cce4
·
verified ·
1 Parent(s): 59eb047

Synchronizing local compiler cache.

Browse files
Files changed (34) hide show
  1. .gitattributes +9 -0
  2. neuronxcc-2.26.6360.0+6f180f47/0_REGISTRY/0.4.7.dev0/719fe7a9c6374369ebb3c8b70e3ca2e20d29e8d4c6f1965131b188cd23d90d00/90a037964593d4d7eed0.json +57 -0
  3. neuronxcc-2.26.6360.0+6f180f47/0_REGISTRY/0.4.7.dev0/929b02754a13cbfdf657d863c3fc6f3bce672879bc6ae48ab45be21e881e9ec2/861439ebd56397783822.json +88 -0
  4. neuronxcc-2.26.6360.0+6f180f47/0_REGISTRY/0.4.7.dev0/c44eac5d03bacf7643a3782a45cab2feff584652bee95cd3c43c7d2600b2f3c5/8452568ed9fe0799584a.json +64 -0
  5. neuronxcc-2.26.6360.0+6f180f47/0_REGISTRY/0.4.7.dev0/llama/TinyLlama/TinyLlama-1.1B-Chat-v1.0/90a037964593d4d7eed0.json +57 -0
  6. neuronxcc-2.26.6360.0+6f180f47/0_REGISTRY/0.4.7.dev0/llama/unsloth/Llama-3.1-8B-Instruct/8452568ed9fe0799584a.json +64 -0
  7. neuronxcc-2.26.6360.0+6f180f47/0_REGISTRY/0.4.7.dev0/qwen3/Qwen/Qwen3-0.6B/861439ebd56397783822.json +88 -0
  8. neuronxcc-2.26.6360.0+6f180f47/MODULE_45c0bed0ef53eb70769d+665810b0/compile_flags.json +1 -0
  9. neuronxcc-2.26.6360.0+6f180f47/MODULE_45c0bed0ef53eb70769d+665810b0/model.done +0 -0
  10. neuronxcc-2.26.6360.0+6f180f47/MODULE_45c0bed0ef53eb70769d+665810b0/model.hlo_module.pb +3 -0
  11. neuronxcc-2.26.6360.0+6f180f47/MODULE_45c0bed0ef53eb70769d+665810b0/model.neff +3 -0
  12. neuronxcc-2.26.6360.0+6f180f47/MODULE_49b256952b6e47c8eee6+665810b0/compile_flags.json +1 -0
  13. neuronxcc-2.26.6360.0+6f180f47/MODULE_49b256952b6e47c8eee6+665810b0/model.done +0 -0
  14. neuronxcc-2.26.6360.0+6f180f47/MODULE_49b256952b6e47c8eee6+665810b0/model.hlo_module.pb +3 -0
  15. neuronxcc-2.26.6360.0+6f180f47/MODULE_49b256952b6e47c8eee6+665810b0/model.neff +3 -0
  16. neuronxcc-2.26.6360.0+6f180f47/MODULE_a361b21eaf4d1aecc789+4a27301b/compile_flags.json +1 -0
  17. neuronxcc-2.26.6360.0+6f180f47/MODULE_a361b21eaf4d1aecc789+4a27301b/model.done +0 -0
  18. neuronxcc-2.26.6360.0+6f180f47/MODULE_a361b21eaf4d1aecc789+4a27301b/model.hlo_module.pb +3 -0
  19. neuronxcc-2.26.6360.0+6f180f47/MODULE_a361b21eaf4d1aecc789+4a27301b/model.neff +3 -0
  20. neuronxcc-2.26.6360.0+6f180f47/MODULE_a361b21eaf4d1aecc789+4a27301b/wrapped_neff.hlo +3 -0
  21. neuronxcc-2.26.6360.0+6f180f47/MODULE_d07583c884496d483ccc+e457fe88/compile_flags.json +1 -0
  22. neuronxcc-2.26.6360.0+6f180f47/MODULE_d07583c884496d483ccc+e457fe88/model.done +0 -0
  23. neuronxcc-2.26.6360.0+6f180f47/MODULE_d07583c884496d483ccc+e457fe88/model.hlo_module.pb +3 -0
  24. neuronxcc-2.26.6360.0+6f180f47/MODULE_d07583c884496d483ccc+e457fe88/model.neff +3 -0
  25. neuronxcc-2.26.6360.0+6f180f47/MODULE_e549b39486d25778527d+4a27301b/compile_flags.json +1 -0
  26. neuronxcc-2.26.6360.0+6f180f47/MODULE_e549b39486d25778527d+4a27301b/model.done +0 -0
  27. neuronxcc-2.26.6360.0+6f180f47/MODULE_e549b39486d25778527d+4a27301b/model.hlo_module.pb +3 -0
  28. neuronxcc-2.26.6360.0+6f180f47/MODULE_e549b39486d25778527d+4a27301b/model.neff +3 -0
  29. neuronxcc-2.26.6360.0+6f180f47/MODULE_e549b39486d25778527d+4a27301b/wrapped_neff.hlo +3 -0
  30. neuronxcc-2.26.6360.0+6f180f47/MODULE_edf426a5de28b5904f80+4a27301b/compile_flags.json +1 -0
  31. neuronxcc-2.26.6360.0+6f180f47/MODULE_edf426a5de28b5904f80+4a27301b/model.done +0 -0
  32. neuronxcc-2.26.6360.0+6f180f47/MODULE_edf426a5de28b5904f80+4a27301b/model.hlo_module.pb +3 -0
  33. neuronxcc-2.26.6360.0+6f180f47/MODULE_edf426a5de28b5904f80+4a27301b/model.neff +3 -0
  34. neuronxcc-2.26.6360.0+6f180f47/MODULE_edf426a5de28b5904f80+4a27301b/wrapped_neff.hlo +3 -0
.gitattributes CHANGED
@@ -8368,3 +8368,12 @@ neuronxcc-2.26.6360.0+6f180f47/MODULE_fee5fa3b97115a883ace+519f203d/model.neff f
8368
  neuronxcc-2.26.6360.0+6f180f47/MODULE_4a0a3172070c277684d6+519f203d/model.neff filter=lfs diff=lfs merge=lfs -text
8369
  neuronxcc-2.26.6360.0+6f180f47/MODULE_f31a9058de7972c77c79+84337dd9/model.neff filter=lfs diff=lfs merge=lfs -text
8370
  neuronxcc-2.26.6360.0+6f180f47/MODULE_f31a9058de7972c77c79+84337dd9/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
 
 
 
 
8368
  neuronxcc-2.26.6360.0+6f180f47/MODULE_4a0a3172070c277684d6+519f203d/model.neff filter=lfs diff=lfs merge=lfs -text
8369
  neuronxcc-2.26.6360.0+6f180f47/MODULE_f31a9058de7972c77c79+84337dd9/model.neff filter=lfs diff=lfs merge=lfs -text
8370
  neuronxcc-2.26.6360.0+6f180f47/MODULE_f31a9058de7972c77c79+84337dd9/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
8371
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_45c0bed0ef53eb70769d+665810b0/model.neff filter=lfs diff=lfs merge=lfs -text
8372
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_49b256952b6e47c8eee6+665810b0/model.neff filter=lfs diff=lfs merge=lfs -text
8373
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_a361b21eaf4d1aecc789+4a27301b/model.neff filter=lfs diff=lfs merge=lfs -text
8374
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_a361b21eaf4d1aecc789+4a27301b/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
8375
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_d07583c884496d483ccc+e457fe88/model.neff filter=lfs diff=lfs merge=lfs -text
8376
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_e549b39486d25778527d+4a27301b/model.neff filter=lfs diff=lfs merge=lfs -text
8377
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_e549b39486d25778527d+4a27301b/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
8378
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_edf426a5de28b5904f80+4a27301b/model.neff filter=lfs diff=lfs merge=lfs -text
8379
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_edf426a5de28b5904f80+4a27301b/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.26.6360.0+6f180f47/0_REGISTRY/0.4.7.dev0/719fe7a9c6374369ebb3c8b70e3ca2e20d29e8d4c6f1965131b188cd23d90d00/90a037964593d4d7eed0.json ADDED
@@ -0,0 +1,57 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "TinyLlama/TinyLlama-1.1B-Chat-v1.0",
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": 5632,
16
+ "max_position_embeddings": 2048,
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": "TinyLlama/TinyLlama-1.1B-Chat-v1.0",
24
+ "checkpoint_revision": "fe8a4ea1ffedaf415f4da2f062534de366a451e6",
25
+ "continuous_batching": false,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 4,
30
+ "max_batch_size": 1,
31
+ "max_context_length": 2048,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 2048,
34
+ "neuronxcc_version": "2.26.6360.0+6f180f47",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.7.dev0",
37
+ "output_logits": false,
38
+ "pp_degree": 1,
39
+ "prefill_chunk_size": 1024,
40
+ "sequence_length": 2048,
41
+ "speculation_length": 0,
42
+ "start_rank_id": 0,
43
+ "target": "trn2",
44
+ "torch_dtype": "bfloat16",
45
+ "tp_degree": 4
46
+ },
47
+ "num_attention_heads": 32,
48
+ "num_hidden_layers": 22,
49
+ "num_key_value_heads": 4,
50
+ "pretraining_tp": 1,
51
+ "rms_norm_eps": 1e-05,
52
+ "rope_scaling": null,
53
+ "rope_theta": 10000.0,
54
+ "tie_word_embeddings": false,
55
+ "use_cache": true,
56
+ "vocab_size": 32000
57
+ }
neuronxcc-2.26.6360.0+6f180f47/0_REGISTRY/0.4.7.dev0/929b02754a13cbfdf657d863c3fc6f3bce672879bc6ae48ab45be21e881e9ec2/861439ebd56397783822.json ADDED
@@ -0,0 +1,88 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": 4,
52
+ "capacity_factor": null,
53
+ "checkpoint_id": "Qwen/Qwen3-0.6B",
54
+ "checkpoint_revision": "c1899de289a04d12100db370d81485cdf75e47ca",
55
+ "continuous_batching": true,
56
+ "ep_degree": 1,
57
+ "fused_qkv": true,
58
+ "glu_mlp": true,
59
+ "local_ranks_size": 1,
60
+ "max_batch_size": 4,
61
+ "max_context_length": 1024,
62
+ "max_topk": 256,
63
+ "n_active_tokens": 1024,
64
+ "neuronxcc_version": "2.26.6360.0+6f180f47",
65
+ "on_device_sampling": true,
66
+ "optimum_neuron_version": "0.4.7.dev0",
67
+ "output_logits": false,
68
+ "pp_degree": 1,
69
+ "prefill_chunk_size": 0,
70
+ "sequence_length": 1024,
71
+ "speculation_length": 0,
72
+ "start_rank_id": 0,
73
+ "target": "trn2",
74
+ "torch_dtype": "bfloat16",
75
+ "tp_degree": 1
76
+ },
77
+ "num_attention_heads": 16,
78
+ "num_hidden_layers": 28,
79
+ "num_key_value_heads": 8,
80
+ "rms_norm_eps": 1e-06,
81
+ "rope_scaling": null,
82
+ "rope_theta": 1000000,
83
+ "sliding_window": null,
84
+ "tie_word_embeddings": true,
85
+ "use_cache": true,
86
+ "use_sliding_window": false,
87
+ "vocab_size": 151936
88
+ }
neuronxcc-2.26.6360.0+6f180f47/0_REGISTRY/0.4.7.dev0/c44eac5d03bacf7643a3782a45cab2feff584652bee95cd3c43c7d2600b2f3c5/8452568ed9fe0799584a.json ADDED
@@ -0,0 +1,64 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "unsloth/Llama-3.1-8B-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": 128,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 4096,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 14336,
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-3.1-8B-Instruct",
24
+ "checkpoint_revision": "4699cc75b550f9c6f3173fb80f4703b62d946aa5",
25
+ "continuous_batching": false,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 1,
30
+ "max_batch_size": 1,
31
+ "max_context_length": 4096,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 4096,
34
+ "neuronxcc_version": "2.26.6360.0+6f180f47",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.7.dev0",
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": "trn2",
44
+ "torch_dtype": "bfloat16",
45
+ "tp_degree": 1
46
+ },
47
+ "num_attention_heads": 32,
48
+ "num_hidden_layers": 32,
49
+ "num_key_value_heads": 8,
50
+ "pretraining_tp": 1,
51
+ "rms_norm_eps": 1e-05,
52
+ "rope_scaling": {
53
+ "factor": 8.0,
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": false,
61
+ "unsloth_fixed": true,
62
+ "use_cache": true,
63
+ "vocab_size": 128256
64
+ }
neuronxcc-2.26.6360.0+6f180f47/0_REGISTRY/0.4.7.dev0/llama/TinyLlama/TinyLlama-1.1B-Chat-v1.0/90a037964593d4d7eed0.json ADDED
@@ -0,0 +1,57 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "TinyLlama/TinyLlama-1.1B-Chat-v1.0",
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": 5632,
16
+ "max_position_embeddings": 2048,
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": "TinyLlama/TinyLlama-1.1B-Chat-v1.0",
24
+ "checkpoint_revision": "fe8a4ea1ffedaf415f4da2f062534de366a451e6",
25
+ "continuous_batching": false,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 4,
30
+ "max_batch_size": 1,
31
+ "max_context_length": 2048,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 2048,
34
+ "neuronxcc_version": "2.26.6360.0+6f180f47",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.7.dev0",
37
+ "output_logits": false,
38
+ "pp_degree": 1,
39
+ "prefill_chunk_size": 1024,
40
+ "sequence_length": 2048,
41
+ "speculation_length": 0,
42
+ "start_rank_id": 0,
43
+ "target": "trn2",
44
+ "torch_dtype": "bfloat16",
45
+ "tp_degree": 4
46
+ },
47
+ "num_attention_heads": 32,
48
+ "num_hidden_layers": 22,
49
+ "num_key_value_heads": 4,
50
+ "pretraining_tp": 1,
51
+ "rms_norm_eps": 1e-05,
52
+ "rope_scaling": null,
53
+ "rope_theta": 10000.0,
54
+ "tie_word_embeddings": false,
55
+ "use_cache": true,
56
+ "vocab_size": 32000
57
+ }
neuronxcc-2.26.6360.0+6f180f47/0_REGISTRY/0.4.7.dev0/llama/unsloth/Llama-3.1-8B-Instruct/8452568ed9fe0799584a.json ADDED
@@ -0,0 +1,64 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "unsloth/Llama-3.1-8B-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": 128,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 4096,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 14336,
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-3.1-8B-Instruct",
24
+ "checkpoint_revision": "4699cc75b550f9c6f3173fb80f4703b62d946aa5",
25
+ "continuous_batching": false,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 1,
30
+ "max_batch_size": 1,
31
+ "max_context_length": 4096,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 4096,
34
+ "neuronxcc_version": "2.26.6360.0+6f180f47",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.7.dev0",
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": "trn2",
44
+ "torch_dtype": "bfloat16",
45
+ "tp_degree": 1
46
+ },
47
+ "num_attention_heads": 32,
48
+ "num_hidden_layers": 32,
49
+ "num_key_value_heads": 8,
50
+ "pretraining_tp": 1,
51
+ "rms_norm_eps": 1e-05,
52
+ "rope_scaling": {
53
+ "factor": 8.0,
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": false,
61
+ "unsloth_fixed": true,
62
+ "use_cache": true,
63
+ "vocab_size": 128256
64
+ }
neuronxcc-2.26.6360.0+6f180f47/0_REGISTRY/0.4.7.dev0/qwen3/Qwen/Qwen3-0.6B/861439ebd56397783822.json ADDED
@@ -0,0 +1,88 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": 4,
52
+ "capacity_factor": null,
53
+ "checkpoint_id": "Qwen/Qwen3-0.6B",
54
+ "checkpoint_revision": "c1899de289a04d12100db370d81485cdf75e47ca",
55
+ "continuous_batching": true,
56
+ "ep_degree": 1,
57
+ "fused_qkv": true,
58
+ "glu_mlp": true,
59
+ "local_ranks_size": 1,
60
+ "max_batch_size": 4,
61
+ "max_context_length": 1024,
62
+ "max_topk": 256,
63
+ "n_active_tokens": 1024,
64
+ "neuronxcc_version": "2.26.6360.0+6f180f47",
65
+ "on_device_sampling": true,
66
+ "optimum_neuron_version": "0.4.7.dev0",
67
+ "output_logits": false,
68
+ "pp_degree": 1,
69
+ "prefill_chunk_size": 0,
70
+ "sequence_length": 1024,
71
+ "speculation_length": 0,
72
+ "start_rank_id": 0,
73
+ "target": "trn2",
74
+ "torch_dtype": "bfloat16",
75
+ "tp_degree": 1
76
+ },
77
+ "num_attention_heads": 16,
78
+ "num_hidden_layers": 28,
79
+ "num_key_value_heads": 8,
80
+ "rms_norm_eps": 1e-06,
81
+ "rope_scaling": null,
82
+ "rope_theta": 1000000,
83
+ "sliding_window": null,
84
+ "tie_word_embeddings": true,
85
+ "use_cache": true,
86
+ "use_sliding_window": false,
87
+ "vocab_size": 151936
88
+ }
neuronxcc-2.26.6360.0+6f180f47/MODULE_45c0bed0ef53eb70769d+665810b0/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn2", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=2", "--verbose=35", "--logfile=/tmp/nxd_model/chunked_prefill/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.26.6360.0+6f180f47/MODULE_45c0bed0ef53eb70769d+665810b0/model.done ADDED
File without changes
neuronxcc-2.26.6360.0+6f180f47/MODULE_45c0bed0ef53eb70769d+665810b0/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:30d901717677b45da4031b4966382ca6dc1388d5ca3f4c51736425de2571b5cb
3
+ size 3864154
neuronxcc-2.26.6360.0+6f180f47/MODULE_45c0bed0ef53eb70769d+665810b0/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9dd2de91c1e05ddfa81c191c4838eb8c88b00858c55d11602393e8a65e10cc76
3
+ size 11822490
neuronxcc-2.26.6360.0+6f180f47/MODULE_49b256952b6e47c8eee6+665810b0/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn2", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=2", "--verbose=35", "--logfile=/tmp/nxd_model/chunked_prefill/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.26.6360.0+6f180f47/MODULE_49b256952b6e47c8eee6+665810b0/model.done ADDED
File without changes
neuronxcc-2.26.6360.0+6f180f47/MODULE_49b256952b6e47c8eee6+665810b0/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9b9672f81b6e5c2c243c2e11778ce8a0a2583a88da0e45a0f92d013a0e191178
3
+ size 2831198
neuronxcc-2.26.6360.0+6f180f47/MODULE_49b256952b6e47c8eee6+665810b0/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4187401636809981d251d583523f42436eea8a193662e22a0f04239806e71b7f
3
+ size 5340742
neuronxcc-2.26.6360.0+6f180f47/MODULE_a361b21eaf4d1aecc789+4a27301b/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn2", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=2", "--verbose=35", "--logfile=/tmp/nxd_model/token_generation/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
neuronxcc-2.26.6360.0+6f180f47/MODULE_a361b21eaf4d1aecc789+4a27301b/model.done ADDED
File without changes
neuronxcc-2.26.6360.0+6f180f47/MODULE_a361b21eaf4d1aecc789+4a27301b/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:05300fecbb9305be6b0b14452ebbe4f064a2ebbccc9724ddaaec7b6a9d5c34ac
3
+ size 2885108
neuronxcc-2.26.6360.0+6f180f47/MODULE_a361b21eaf4d1aecc789+4a27301b/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:48e608b17f04dfd8416b80d0c02935de810192107ed409f34b2cd86ee2c13284
3
+ size 18730935
neuronxcc-2.26.6360.0+6f180f47/MODULE_a361b21eaf4d1aecc789+4a27301b/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ead3c445372e1d69642e123c974cec8eceb62612b43167219326cea4fb12afec
3
+ size 18881118
neuronxcc-2.26.6360.0+6f180f47/MODULE_d07583c884496d483ccc+e457fe88/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn2", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=2", "--verbose=35", "--logfile=/tmp/nxd_model/context_encoding/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.26.6360.0+6f180f47/MODULE_d07583c884496d483ccc+e457fe88/model.done ADDED
File without changes
neuronxcc-2.26.6360.0+6f180f47/MODULE_d07583c884496d483ccc+e457fe88/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:245952864b8a3d8b5ccd623e7180cdb6ef98171429f62202fa8791117671bda6
3
+ size 942150
neuronxcc-2.26.6360.0+6f180f47/MODULE_d07583c884496d483ccc+e457fe88/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7b575513f3db827eb5a608e188f317267f6865da59981997c880200b5e509d0b
3
+ size 6099547
neuronxcc-2.26.6360.0+6f180f47/MODULE_e549b39486d25778527d+4a27301b/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn2", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=2", "--verbose=35", "--logfile=/tmp/nxd_model/token_generation/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
neuronxcc-2.26.6360.0+6f180f47/MODULE_e549b39486d25778527d+4a27301b/model.done ADDED
File without changes
neuronxcc-2.26.6360.0+6f180f47/MODULE_e549b39486d25778527d+4a27301b/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:46dea878124ca60699796c4ce53cada51d808c394fe53f82a3b36d6553d79321
3
+ size 2167725
neuronxcc-2.26.6360.0+6f180f47/MODULE_e549b39486d25778527d+4a27301b/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3fe73a264e3f5fedcb0181cf564991f265d6ad8668a8ae1c70070101f9afc463
3
+ size 1599979
neuronxcc-2.26.6360.0+6f180f47/MODULE_e549b39486d25778527d+4a27301b/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8a5258b70c6fb1805a87c5471f19635597eeecf5aaf323a3689a4958017c6a23
3
+ size 1701504
neuronxcc-2.26.6360.0+6f180f47/MODULE_edf426a5de28b5904f80+4a27301b/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn2", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=2", "--verbose=35", "--logfile=/tmp/nxd_model/token_generation/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
neuronxcc-2.26.6360.0+6f180f47/MODULE_edf426a5de28b5904f80+4a27301b/model.done ADDED
File without changes
neuronxcc-2.26.6360.0+6f180f47/MODULE_edf426a5de28b5904f80+4a27301b/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d27bbfb1db21623864f7f4c6f3fde673d5169d1378a18bcfebcd9a25d6255a6c
3
+ size 683058
neuronxcc-2.26.6360.0+6f180f47/MODULE_edf426a5de28b5904f80+4a27301b/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:635b3d0a1e9f9210b8459e09664a95816f0b3b9559c03e81fd387c75846c190b
3
+ size 4809190
neuronxcc-2.26.6360.0+6f180f47/MODULE_edf426a5de28b5904f80+4a27301b/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:17be33798a73809eb1eb654d2ccac762482354bf4ebfd35068411461b3a866ff
3
+ size 4947299