Hugging Face's logo

optimum-internal-testing
/
neuron-testing-cache

Model card Files
xet
Community
dacorvo HF Staff commited on
Commit
4fabdab
·
verified ·
1 Parent(s): 377b572

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 +16 -0
  2. neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev0/llama/meta-llama/Llama-3.2-1B/71d2d9a950004b288ebe.json +77 -0
  3. neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev0/qwen2/yujiepan/qwen2.5-128k-tiny-random/3292185f0123466a93a2.json +79 -0
  4. neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev0/qwen2/yujiepan/qwen2.5-128k-tiny-random/b60dca7d7a9c53810fe0.json +79 -0
  5. neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev0/qwen2/yujiepan/qwen2.5-128k-tiny-random/b9b0c081795fa9e33ca6.json +79 -0
  6. neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev0/smollm3/HuggingFaceTB/SmolLM3-3B/5967f2b3d54a7c5a3168.json +149 -0
  7. neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev0/smollm3/HuggingFaceTB/SmolLM3-3B/c8a379bab933fc9c554c.json +149 -0
  8. neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev0/smollm3/HuggingFaceTB/SmolLM3-3B/d76398f17f64a0ddfb0a.json +149 -0
  9. neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/granite/hf-internal-testing/tiny-random-GraniteForCausalLM/60271363e0e147450b0f.json +73 -0
  10. neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/granite/hf-internal-testing/tiny-random-GraniteForCausalLM/a9e5f04b708cd397fbdd.json +73 -0
  11. neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/granite/hf-internal-testing/tiny-random-GraniteForCausalLM/de79a3968941a6d3864f.json +73 -0
  12. neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/llama/llamafactory/tiny-random-Llama-3/28b90633a77c30e8a187.json +77 -0
  13. neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/llama/llamafactory/tiny-random-Llama-3/88a97541d018d7aa0b29.json +77 -0
  14. neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/llama/llamafactory/tiny-random-Llama-3/d8f338a72f4bb0a4682d.json +77 -0
  15. neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/llama/unsloth/Llama-3.2-1B-Instruct/f7ed18f9ae29fab731ad.json +78 -0
  16. neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/phi3/yujiepan/phi-4-tiny-random/0d648a236d7cc578c5a8.json +74 -0
  17. neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/phi3/yujiepan/phi-4-tiny-random/1228009b3909ec8d9751.json +74 -0
  18. neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/phi3/yujiepan/phi-4-tiny-random/1c5104d89e025d3fb8db.json +74 -0
  19. neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/qwen2/yujiepan/qwen2.5-128k-tiny-random/4f65451820dd0eadcdca.json +79 -0
  20. neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/qwen2/yujiepan/qwen2.5-128k-tiny-random/d7566c1b656e75712c0c.json +79 -0
  21. neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/qwen2/yujiepan/qwen2.5-128k-tiny-random/f47233e40681eaec5807.json +79 -0
  22. neuronxcc-2.19.8089.0+8ab9f450/MODULE_018530e1260c6c7cee60+ed72d204/compile_flags.json +1 -0
  23. neuronxcc-2.19.8089.0+8ab9f450/MODULE_018530e1260c6c7cee60+ed72d204/model.done +0 -0
  24. neuronxcc-2.19.8089.0+8ab9f450/MODULE_018530e1260c6c7cee60+ed72d204/model.hlo_module.pb +3 -0
  25. neuronxcc-2.19.8089.0+8ab9f450/MODULE_018530e1260c6c7cee60+ed72d204/model.neff +3 -0
  26. neuronxcc-2.19.8089.0+8ab9f450/MODULE_16dadb9b6ab4bb044c1e+a9d440f5/compile_flags.json +1 -0
  27. neuronxcc-2.19.8089.0+8ab9f450/MODULE_16dadb9b6ab4bb044c1e+a9d440f5/model.done +0 -0
  28. neuronxcc-2.19.8089.0+8ab9f450/MODULE_16dadb9b6ab4bb044c1e+a9d440f5/model.hlo_module.pb +3 -0
  29. neuronxcc-2.19.8089.0+8ab9f450/MODULE_16dadb9b6ab4bb044c1e+a9d440f5/model.neff +3 -0
  30. neuronxcc-2.19.8089.0+8ab9f450/MODULE_16dadb9b6ab4bb044c1e+a9d440f5/wrapped_neff.hlo +3 -0
  31. neuronxcc-2.19.8089.0+8ab9f450/MODULE_1d2b7dee00e90e52700e+431f5505/compile_flags.json +1 -0
  32. neuronxcc-2.19.8089.0+8ab9f450/MODULE_1d2b7dee00e90e52700e+431f5505/model.done +0 -0
  33. neuronxcc-2.19.8089.0+8ab9f450/MODULE_1d2b7dee00e90e52700e+431f5505/model.hlo_module.pb +3 -0
  34. neuronxcc-2.19.8089.0+8ab9f450/MODULE_1d2b7dee00e90e52700e+431f5505/model.neff +3 -0
  35. neuronxcc-2.19.8089.0+8ab9f450/MODULE_41d952393bdf990574ea+431f5505/compile_flags.json +1 -0
  36. neuronxcc-2.19.8089.0+8ab9f450/MODULE_41d952393bdf990574ea+431f5505/model.done +0 -0
  37. neuronxcc-2.19.8089.0+8ab9f450/MODULE_41d952393bdf990574ea+431f5505/model.hlo_module.pb +3 -0
  38. neuronxcc-2.19.8089.0+8ab9f450/MODULE_41d952393bdf990574ea+431f5505/model.neff +3 -0
  39. neuronxcc-2.19.8089.0+8ab9f450/MODULE_42f2f26b5d67b722528c+431f5505/compile_flags.json +1 -0
  40. neuronxcc-2.19.8089.0+8ab9f450/MODULE_42f2f26b5d67b722528c+431f5505/model.done +0 -0
  41. neuronxcc-2.19.8089.0+8ab9f450/MODULE_42f2f26b5d67b722528c+431f5505/model.hlo_module.pb +3 -0
  42. neuronxcc-2.19.8089.0+8ab9f450/MODULE_42f2f26b5d67b722528c+431f5505/model.neff +3 -0
  43. neuronxcc-2.19.8089.0+8ab9f450/MODULE_5a4f005f17fc9cd3a4ea+a9d440f5/compile_flags.json +1 -0
  44. neuronxcc-2.19.8089.0+8ab9f450/MODULE_5a4f005f17fc9cd3a4ea+a9d440f5/model.done +0 -0
  45. neuronxcc-2.19.8089.0+8ab9f450/MODULE_5a4f005f17fc9cd3a4ea+a9d440f5/model.hlo_module.pb +3 -0
  46. neuronxcc-2.19.8089.0+8ab9f450/MODULE_5a4f005f17fc9cd3a4ea+a9d440f5/model.neff +3 -0
  47. neuronxcc-2.19.8089.0+8ab9f450/MODULE_5a4f005f17fc9cd3a4ea+a9d440f5/wrapped_neff.hlo +3 -0
  48. neuronxcc-2.19.8089.0+8ab9f450/MODULE_865d84616d5b312d9a15+ed72d204/compile_flags.json +1 -0
  49. neuronxcc-2.19.8089.0+8ab9f450/MODULE_865d84616d5b312d9a15+ed72d204/model.done +0 -0
  50. neuronxcc-2.19.8089.0+8ab9f450/MODULE_865d84616d5b312d9a15+ed72d204/model.hlo_module.pb +3 -0
.gitattributes CHANGED
@@ -3804,3 +3804,19 @@ neuronxcc-2.19.8089.0+8ab9f450/MODULE_91501952fa085a11a635+ed72d204/model.neff f
3804
  neuronxcc-2.19.8089.0+8ab9f450/MODULE_efca9e0e049746cedc78+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
3805
  neuronxcc-2.19.8089.0+8ab9f450/MODULE_efca9e0e049746cedc78+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
3806
  neuronxcc-2.19.8089.0+8ab9f450/MODULE_f5968e810b1a450b5c85+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3804
  neuronxcc-2.19.8089.0+8ab9f450/MODULE_efca9e0e049746cedc78+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
3805
  neuronxcc-2.19.8089.0+8ab9f450/MODULE_efca9e0e049746cedc78+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
3806
  neuronxcc-2.19.8089.0+8ab9f450/MODULE_f5968e810b1a450b5c85+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
3807
+ neuronxcc-2.19.8089.0+8ab9f450/MODULE_018530e1260c6c7cee60+ed72d204/model.neff filter=lfs diff=lfs merge=lfs -text
3808
+ neuronxcc-2.19.8089.0+8ab9f450/MODULE_16dadb9b6ab4bb044c1e+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
3809
+ neuronxcc-2.19.8089.0+8ab9f450/MODULE_16dadb9b6ab4bb044c1e+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
3810
+ neuronxcc-2.19.8089.0+8ab9f450/MODULE_1d2b7dee00e90e52700e+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
3811
+ neuronxcc-2.19.8089.0+8ab9f450/MODULE_41d952393bdf990574ea+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
3812
+ neuronxcc-2.19.8089.0+8ab9f450/MODULE_42f2f26b5d67b722528c+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
3813
+ neuronxcc-2.19.8089.0+8ab9f450/MODULE_5a4f005f17fc9cd3a4ea+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
3814
+ neuronxcc-2.19.8089.0+8ab9f450/MODULE_5a4f005f17fc9cd3a4ea+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
3815
+ neuronxcc-2.19.8089.0+8ab9f450/MODULE_865d84616d5b312d9a15+ed72d204/model.neff filter=lfs diff=lfs merge=lfs -text
3816
+ neuronxcc-2.19.8089.0+8ab9f450/MODULE_93f2290cdf5486e6695f+ed72d204/model.neff filter=lfs diff=lfs merge=lfs -text
3817
+ neuronxcc-2.19.8089.0+8ab9f450/MODULE_a58dd9102c0f9872318f+ed72d204/model.neff filter=lfs diff=lfs merge=lfs -text
3818
+ neuronxcc-2.19.8089.0+8ab9f450/MODULE_c42e5a3d424f33bed45c+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
3819
+ neuronxcc-2.19.8089.0+8ab9f450/MODULE_c42e5a3d424f33bed45c+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
3820
+ neuronxcc-2.19.8089.0+8ab9f450/MODULE_e2af0f8dc3b0905ea1be+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
3821
+ neuronxcc-2.19.8089.0+8ab9f450/MODULE_e2af0f8dc3b0905ea1be+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
3822
+ neuronxcc-2.19.8089.0+8ab9f450/MODULE_fb00ac5d5c23637419fa+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev0/llama/meta-llama/Llama-3.2-1B/71d2d9a950004b288ebe.json ADDED
@@ -0,0 +1,77 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "meta-llama/Llama-3.2-1B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "head_dim": 64,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 2048,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 8192,
15
+ "max_position_embeddings": 131072,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "neuron": {
19
+ "_serialized_key": "NxDNeuronConfig",
20
+ "async_mode": false,
21
+ "attn_kernel_enabled": false,
22
+ "batch_size": 1,
23
+ "capacity_factor": null,
24
+ "cc_pipeline_tiling_factor": 2,
25
+ "checkpoint_id": "meta-llama/Llama-3.2-1B",
26
+ "checkpoint_revision": "4e20de362430cd3b72f300e6b0f18e50e7166e08",
27
+ "continuous_batching": false,
28
+ "enable_bucketing": false,
29
+ "ep_degree": 1,
30
+ "flash_decoding_enabled": false,
31
+ "fused_qkv": true,
32
+ "glu_mlp": true,
33
+ "is_chunked_prefill": false,
34
+ "local_ranks_size": 2,
35
+ "logical_nc_config": 1,
36
+ "max_batch_size": 1,
37
+ "max_context_length": 4096,
38
+ "max_topk": 256,
39
+ "mlp_kernel_enabled": false,
40
+ "mlp_kernel_fuse_residual_add": false,
41
+ "n_active_tokens": 4096,
42
+ "neuronxcc_version": "2.19.8089.0+8ab9f450",
43
+ "num_cores_per_group": 1,
44
+ "on_device_sampling": true,
45
+ "optimum_neuron_version": "0.3.1.dev0",
46
+ "output_logits": false,
47
+ "padding_side": "right",
48
+ "pp_degree": 1,
49
+ "qk_layernorm": false,
50
+ "qkv_kernel_enabled": false,
51
+ "rpl_reduce_dtype": "bfloat16",
52
+ "sequence_length": 4096,
53
+ "sequence_parallel_enabled": false,
54
+ "speculation_length": 0,
55
+ "start_rank_id": 0,
56
+ "target": null,
57
+ "torch_dtype": "bfloat16",
58
+ "tp_degree": 2,
59
+ "vocab_parallel": false
60
+ },
61
+ "num_attention_heads": 32,
62
+ "num_hidden_layers": 16,
63
+ "num_key_value_heads": 8,
64
+ "pretraining_tp": 1,
65
+ "rms_norm_eps": 1e-05,
66
+ "rope_scaling": {
67
+ "factor": 32.0,
68
+ "high_freq_factor": 4.0,
69
+ "low_freq_factor": 1.0,
70
+ "original_max_position_embeddings": 8192,
71
+ "rope_type": "llama3"
72
+ },
73
+ "rope_theta": 500000.0,
74
+ "tie_word_embeddings": true,
75
+ "use_cache": true,
76
+ "vocab_size": 128256
77
+ }
neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev0/qwen2/yujiepan/qwen2.5-128k-tiny-random/3292185f0123466a93a2.json ADDED
@@ -0,0 +1,79 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "yujiepan/qwen2.5-128k-tiny-random",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Qwen2ForCausalLM"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "hidden_act": "silu",
10
+ "hidden_size": 8,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 16,
13
+ "layer_types": [
14
+ "full_attention",
15
+ "full_attention"
16
+ ],
17
+ "max_position_embeddings": 32768,
18
+ "max_window_layers": 1,
19
+ "model_type": "qwen2",
20
+ "neuron": {
21
+ "_serialized_key": "NxDNeuronConfig",
22
+ "async_mode": false,
23
+ "attn_kernel_enabled": false,
24
+ "batch_size": 2,
25
+ "capacity_factor": null,
26
+ "cc_pipeline_tiling_factor": 2,
27
+ "checkpoint_id": "yujiepan/qwen2.5-128k-tiny-random",
28
+ "checkpoint_revision": "c8296d4ca3f87782876d2382fbb6481d1beb8ef0",
29
+ "continuous_batching": true,
30
+ "enable_bucketing": false,
31
+ "ep_degree": 1,
32
+ "flash_decoding_enabled": false,
33
+ "fused_qkv": false,
34
+ "glu_mlp": true,
35
+ "is_chunked_prefill": false,
36
+ "local_ranks_size": 2,
37
+ "logical_nc_config": 1,
38
+ "max_batch_size": 2,
39
+ "max_context_length": 100,
40
+ "max_topk": 256,
41
+ "mlp_kernel_enabled": false,
42
+ "mlp_kernel_fuse_residual_add": false,
43
+ "n_active_tokens": 100,
44
+ "neuronxcc_version": "2.19.8089.0+8ab9f450",
45
+ "num_cores_per_group": 1,
46
+ "on_device_sampling": false,
47
+ "optimum_neuron_version": "0.3.1.dev0",
48
+ "output_logits": false,
49
+ "padding_side": "right",
50
+ "pp_degree": 1,
51
+ "qk_layernorm": false,
52
+ "qkv_kernel_enabled": false,
53
+ "rpl_reduce_dtype": "float16",
54
+ "sequence_length": 100,
55
+ "sequence_parallel_enabled": false,
56
+ "speculation_length": 0,
57
+ "start_rank_id": 0,
58
+ "target": null,
59
+ "torch_dtype": "float16",
60
+ "tp_degree": 2,
61
+ "vocab_parallel": false
62
+ },
63
+ "num_attention_heads": 4,
64
+ "num_hidden_layers": 2,
65
+ "num_key_value_heads": 2,
66
+ "rms_norm_eps": 1e-06,
67
+ "rope_scaling": {
68
+ "factor": 4.0,
69
+ "original_max_position_embeddings": 32768,
70
+ "rope_type": "yarn",
71
+ "type": "yarn"
72
+ },
73
+ "rope_theta": 1000000.0,
74
+ "sliding_window": null,
75
+ "tie_word_embeddings": false,
76
+ "use_cache": true,
77
+ "use_sliding_window": false,
78
+ "vocab_size": 152064
79
+ }
neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev0/qwen2/yujiepan/qwen2.5-128k-tiny-random/b60dca7d7a9c53810fe0.json ADDED
@@ -0,0 +1,79 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "yujiepan/qwen2.5-128k-tiny-random",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Qwen2ForCausalLM"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "hidden_act": "silu",
10
+ "hidden_size": 8,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 16,
13
+ "layer_types": [
14
+ "full_attention",
15
+ "full_attention"
16
+ ],
17
+ "max_position_embeddings": 32768,
18
+ "max_window_layers": 1,
19
+ "model_type": "qwen2",
20
+ "neuron": {
21
+ "_serialized_key": "NxDNeuronConfig",
22
+ "async_mode": false,
23
+ "attn_kernel_enabled": false,
24
+ "batch_size": 1,
25
+ "capacity_factor": null,
26
+ "cc_pipeline_tiling_factor": 2,
27
+ "checkpoint_id": "yujiepan/qwen2.5-128k-tiny-random",
28
+ "checkpoint_revision": "c8296d4ca3f87782876d2382fbb6481d1beb8ef0",
29
+ "continuous_batching": false,
30
+ "enable_bucketing": false,
31
+ "ep_degree": 1,
32
+ "flash_decoding_enabled": false,
33
+ "fused_qkv": false,
34
+ "glu_mlp": true,
35
+ "is_chunked_prefill": false,
36
+ "local_ranks_size": 2,
37
+ "logical_nc_config": 1,
38
+ "max_batch_size": 1,
39
+ "max_context_length": 100,
40
+ "max_topk": 256,
41
+ "mlp_kernel_enabled": false,
42
+ "mlp_kernel_fuse_residual_add": false,
43
+ "n_active_tokens": 100,
44
+ "neuronxcc_version": "2.19.8089.0+8ab9f450",
45
+ "num_cores_per_group": 1,
46
+ "on_device_sampling": true,
47
+ "optimum_neuron_version": "0.3.1.dev0",
48
+ "output_logits": false,
49
+ "padding_side": "right",
50
+ "pp_degree": 1,
51
+ "qk_layernorm": false,
52
+ "qkv_kernel_enabled": false,
53
+ "rpl_reduce_dtype": "float16",
54
+ "sequence_length": 100,
55
+ "sequence_parallel_enabled": false,
56
+ "speculation_length": 0,
57
+ "start_rank_id": 0,
58
+ "target": null,
59
+ "torch_dtype": "float16",
60
+ "tp_degree": 2,
61
+ "vocab_parallel": false
62
+ },
63
+ "num_attention_heads": 4,
64
+ "num_hidden_layers": 2,
65
+ "num_key_value_heads": 2,
66
+ "rms_norm_eps": 1e-06,
67
+ "rope_scaling": {
68
+ "factor": 4.0,
69
+ "original_max_position_embeddings": 32768,
70
+ "rope_type": "yarn",
71
+ "type": "yarn"
72
+ },
73
+ "rope_theta": 1000000.0,
74
+ "sliding_window": null,
75
+ "tie_word_embeddings": false,
76
+ "use_cache": true,
77
+ "use_sliding_window": false,
78
+ "vocab_size": 152064
79
+ }
neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev0/qwen2/yujiepan/qwen2.5-128k-tiny-random/b9b0c081795fa9e33ca6.json ADDED
@@ -0,0 +1,79 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "yujiepan/qwen2.5-128k-tiny-random",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Qwen2ForCausalLM"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "hidden_act": "silu",
10
+ "hidden_size": 8,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 16,
13
+ "layer_types": [
14
+ "full_attention",
15
+ "full_attention"
16
+ ],
17
+ "max_position_embeddings": 32768,
18
+ "max_window_layers": 1,
19
+ "model_type": "qwen2",
20
+ "neuron": {
21
+ "_serialized_key": "NxDNeuronConfig",
22
+ "async_mode": false,
23
+ "attn_kernel_enabled": false,
24
+ "batch_size": 1,
25
+ "capacity_factor": null,
26
+ "cc_pipeline_tiling_factor": 2,
27
+ "checkpoint_id": "yujiepan/qwen2.5-128k-tiny-random",
28
+ "checkpoint_revision": "c8296d4ca3f87782876d2382fbb6481d1beb8ef0",
29
+ "continuous_batching": false,
30
+ "enable_bucketing": false,
31
+ "ep_degree": 1,
32
+ "flash_decoding_enabled": false,
33
+ "fused_qkv": false,
34
+ "glu_mlp": true,
35
+ "is_chunked_prefill": false,
36
+ "local_ranks_size": 2,
37
+ "logical_nc_config": 1,
38
+ "max_batch_size": 1,
39
+ "max_context_length": 100,
40
+ "max_topk": 256,
41
+ "mlp_kernel_enabled": false,
42
+ "mlp_kernel_fuse_residual_add": false,
43
+ "n_active_tokens": 100,
44
+ "neuronxcc_version": "2.19.8089.0+8ab9f450",
45
+ "num_cores_per_group": 1,
46
+ "on_device_sampling": true,
47
+ "optimum_neuron_version": "0.3.1.dev0",
48
+ "output_logits": false,
49
+ "padding_side": "right",
50
+ "pp_degree": 1,
51
+ "qk_layernorm": false,
52
+ "qkv_kernel_enabled": false,
53
+ "rpl_reduce_dtype": "bfloat16",
54
+ "sequence_length": 100,
55
+ "sequence_parallel_enabled": false,
56
+ "speculation_length": 0,
57
+ "start_rank_id": 0,
58
+ "target": null,
59
+ "torch_dtype": "bfloat16",
60
+ "tp_degree": 2,
61
+ "vocab_parallel": false
62
+ },
63
+ "num_attention_heads": 4,
64
+ "num_hidden_layers": 2,
65
+ "num_key_value_heads": 2,
66
+ "rms_norm_eps": 1e-06,
67
+ "rope_scaling": {
68
+ "factor": 4.0,
69
+ "original_max_position_embeddings": 32768,
70
+ "rope_type": "yarn",
71
+ "type": "yarn"
72
+ },
73
+ "rope_theta": 1000000.0,
74
+ "sliding_window": null,
75
+ "tie_word_embeddings": false,
76
+ "use_cache": true,
77
+ "use_sliding_window": false,
78
+ "vocab_size": 152064
79
+ }
neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev0/smollm3/HuggingFaceTB/SmolLM3-3B/5967f2b3d54a7c5a3168.json ADDED
@@ -0,0 +1,149 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "hidden_act": "silu",
11
+ "hidden_size": 2048,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 11008,
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
+ "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
+ ],
52
+ "max_position_embeddings": 65536,
53
+ "max_window_layers": 28,
54
+ "mlp_bias": false,
55
+ "model_type": "smollm3",
56
+ "neuron": {
57
+ "_serialized_key": "NxDNeuronConfig",
58
+ "async_mode": false,
59
+ "attn_kernel_enabled": false,
60
+ "batch_size": 1,
61
+ "capacity_factor": null,
62
+ "cc_pipeline_tiling_factor": 2,
63
+ "checkpoint_id": "HuggingFaceTB/SmolLM3-3B",
64
+ "checkpoint_revision": "988f1dde6bd2a36e3e7f45777a36507ea004fe22",
65
+ "continuous_batching": false,
66
+ "enable_bucketing": false,
67
+ "ep_degree": 1,
68
+ "flash_decoding_enabled": false,
69
+ "fused_qkv": true,
70
+ "glu_mlp": true,
71
+ "is_chunked_prefill": false,
72
+ "local_ranks_size": 2,
73
+ "logical_nc_config": 1,
74
+ "max_batch_size": 1,
75
+ "max_context_length": 4096,
76
+ "max_topk": 256,
77
+ "mlp_kernel_enabled": false,
78
+ "mlp_kernel_fuse_residual_add": false,
79
+ "n_active_tokens": 4096,
80
+ "neuronxcc_version": "2.19.8089.0+8ab9f450",
81
+ "num_cores_per_group": 1,
82
+ "on_device_sampling": false,
83
+ "optimum_neuron_version": "0.3.1.dev0",
84
+ "output_logits": false,
85
+ "padding_side": "right",
86
+ "pp_degree": 1,
87
+ "qk_layernorm": false,
88
+ "qkv_kernel_enabled": false,
89
+ "rpl_reduce_dtype": "bfloat16",
90
+ "sequence_length": 4096,
91
+ "sequence_parallel_enabled": false,
92
+ "speculation_length": 0,
93
+ "start_rank_id": 0,
94
+ "target": null,
95
+ "torch_dtype": "bfloat16",
96
+ "tp_degree": 2,
97
+ "vocab_parallel": false
98
+ },
99
+ "no_rope_layer_interval": 4,
100
+ "no_rope_layers": [
101
+ 1,
102
+ 1,
103
+ 1,
104
+ 0,
105
+ 1,
106
+ 1,
107
+ 1,
108
+ 0,
109
+ 1,
110
+ 1,
111
+ 1,
112
+ 0,
113
+ 1,
114
+ 1,
115
+ 1,
116
+ 0,
117
+ 1,
118
+ 1,
119
+ 1,
120
+ 0,
121
+ 1,
122
+ 1,
123
+ 1,
124
+ 0,
125
+ 1,
126
+ 1,
127
+ 1,
128
+ 0,
129
+ 1,
130
+ 1,
131
+ 1,
132
+ 0,
133
+ 1,
134
+ 1,
135
+ 1,
136
+ 0
137
+ ],
138
+ "num_attention_heads": 16,
139
+ "num_hidden_layers": 36,
140
+ "num_key_value_heads": 4,
141
+ "pretraining_tp": 2,
142
+ "rms_norm_eps": 1e-06,
143
+ "rope_scaling": null,
144
+ "rope_theta": 5000000.0,
145
+ "sliding_window": null,
146
+ "use_cache": false,
147
+ "use_sliding_window": false,
148
+ "vocab_size": 128256
149
+ }
neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev0/smollm3/HuggingFaceTB/SmolLM3-3B/c8a379bab933fc9c554c.json ADDED
@@ -0,0 +1,149 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "hidden_act": "silu",
11
+ "hidden_size": 2048,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 11008,
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
+ "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
+ ],
52
+ "max_position_embeddings": 65536,
53
+ "max_window_layers": 28,
54
+ "mlp_bias": false,
55
+ "model_type": "smollm3",
56
+ "neuron": {
57
+ "_serialized_key": "NxDNeuronConfig",
58
+ "async_mode": false,
59
+ "attn_kernel_enabled": false,
60
+ "batch_size": 1,
61
+ "capacity_factor": null,
62
+ "cc_pipeline_tiling_factor": 2,
63
+ "checkpoint_id": "HuggingFaceTB/SmolLM3-3B",
64
+ "checkpoint_revision": "988f1dde6bd2a36e3e7f45777a36507ea004fe22",
65
+ "continuous_batching": false,
66
+ "enable_bucketing": false,
67
+ "ep_degree": 1,
68
+ "flash_decoding_enabled": false,
69
+ "fused_qkv": false,
70
+ "glu_mlp": true,
71
+ "is_chunked_prefill": false,
72
+ "local_ranks_size": 2,
73
+ "logical_nc_config": 1,
74
+ "max_batch_size": 1,
75
+ "max_context_length": 4096,
76
+ "max_topk": 256,
77
+ "mlp_kernel_enabled": false,
78
+ "mlp_kernel_fuse_residual_add": false,
79
+ "n_active_tokens": 4096,
80
+ "neuronxcc_version": "2.19.8089.0+8ab9f450",
81
+ "num_cores_per_group": 1,
82
+ "on_device_sampling": true,
83
+ "optimum_neuron_version": "0.3.1.dev0",
84
+ "output_logits": false,
85
+ "padding_side": "right",
86
+ "pp_degree": 1,
87
+ "qk_layernorm": false,
88
+ "qkv_kernel_enabled": false,
89
+ "rpl_reduce_dtype": "bfloat16",
90
+ "sequence_length": 4096,
91
+ "sequence_parallel_enabled": false,
92
+ "speculation_length": 0,
93
+ "start_rank_id": 0,
94
+ "target": null,
95
+ "torch_dtype": "bfloat16",
96
+ "tp_degree": 2,
97
+ "vocab_parallel": false
98
+ },
99
+ "no_rope_layer_interval": 4,
100
+ "no_rope_layers": [
101
+ 1,
102
+ 1,
103
+ 1,
104
+ 0,
105
+ 1,
106
+ 1,
107
+ 1,
108
+ 0,
109
+ 1,
110
+ 1,
111
+ 1,
112
+ 0,
113
+ 1,
114
+ 1,
115
+ 1,
116
+ 0,
117
+ 1,
118
+ 1,
119
+ 1,
120
+ 0,
121
+ 1,
122
+ 1,
123
+ 1,
124
+ 0,
125
+ 1,
126
+ 1,
127
+ 1,
128
+ 0,
129
+ 1,
130
+ 1,
131
+ 1,
132
+ 0,
133
+ 1,
134
+ 1,
135
+ 1,
136
+ 0
137
+ ],
138
+ "num_attention_heads": 16,
139
+ "num_hidden_layers": 36,
140
+ "num_key_value_heads": 4,
141
+ "pretraining_tp": 2,
142
+ "rms_norm_eps": 1e-06,
143
+ "rope_scaling": null,
144
+ "rope_theta": 5000000.0,
145
+ "sliding_window": null,
146
+ "use_cache": false,
147
+ "use_sliding_window": false,
148
+ "vocab_size": 128256
149
+ }
neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev0/smollm3/HuggingFaceTB/SmolLM3-3B/d76398f17f64a0ddfb0a.json ADDED
@@ -0,0 +1,149 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "hidden_act": "silu",
11
+ "hidden_size": 2048,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 11008,
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
+ "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
+ ],
52
+ "max_position_embeddings": 65536,
53
+ "max_window_layers": 28,
54
+ "mlp_bias": false,
55
+ "model_type": "smollm3",
56
+ "neuron": {
57
+ "_serialized_key": "NxDNeuronConfig",
58
+ "async_mode": false,
59
+ "attn_kernel_enabled": false,
60
+ "batch_size": 1,
61
+ "capacity_factor": null,
62
+ "cc_pipeline_tiling_factor": 2,
63
+ "checkpoint_id": "HuggingFaceTB/SmolLM3-3B",
64
+ "checkpoint_revision": "988f1dde6bd2a36e3e7f45777a36507ea004fe22",
65
+ "continuous_batching": false,
66
+ "enable_bucketing": false,
67
+ "ep_degree": 1,
68
+ "flash_decoding_enabled": false,
69
+ "fused_qkv": true,
70
+ "glu_mlp": true,
71
+ "is_chunked_prefill": false,
72
+ "local_ranks_size": 2,
73
+ "logical_nc_config": 1,
74
+ "max_batch_size": 1,
75
+ "max_context_length": 4096,
76
+ "max_topk": 256,
77
+ "mlp_kernel_enabled": false,
78
+ "mlp_kernel_fuse_residual_add": false,
79
+ "n_active_tokens": 4096,
80
+ "neuronxcc_version": "2.19.8089.0+8ab9f450",
81
+ "num_cores_per_group": 1,
82
+ "on_device_sampling": true,
83
+ "optimum_neuron_version": "0.3.1.dev0",
84
+ "output_logits": false,
85
+ "padding_side": "right",
86
+ "pp_degree": 1,
87
+ "qk_layernorm": false,
88
+ "qkv_kernel_enabled": false,
89
+ "rpl_reduce_dtype": "bfloat16",
90
+ "sequence_length": 4096,
91
+ "sequence_parallel_enabled": false,
92
+ "speculation_length": 0,
93
+ "start_rank_id": 0,
94
+ "target": null,
95
+ "torch_dtype": "bfloat16",
96
+ "tp_degree": 2,
97
+ "vocab_parallel": false
98
+ },
99
+ "no_rope_layer_interval": 4,
100
+ "no_rope_layers": [
101
+ 1,
102
+ 1,
103
+ 1,
104
+ 0,
105
+ 1,
106
+ 1,
107
+ 1,
108
+ 0,
109
+ 1,
110
+ 1,
111
+ 1,
112
+ 0,
113
+ 1,
114
+ 1,
115
+ 1,
116
+ 0,
117
+ 1,
118
+ 1,
119
+ 1,
120
+ 0,
121
+ 1,
122
+ 1,
123
+ 1,
124
+ 0,
125
+ 1,
126
+ 1,
127
+ 1,
128
+ 0,
129
+ 1,
130
+ 1,
131
+ 1,
132
+ 0,
133
+ 1,
134
+ 1,
135
+ 1,
136
+ 0
137
+ ],
138
+ "num_attention_heads": 16,
139
+ "num_hidden_layers": 36,
140
+ "num_key_value_heads": 4,
141
+ "pretraining_tp": 2,
142
+ "rms_norm_eps": 1e-06,
143
+ "rope_scaling": null,
144
+ "rope_theta": 5000000.0,
145
+ "sliding_window": null,
146
+ "use_cache": false,
147
+ "use_sliding_window": false,
148
+ "vocab_size": 128256
149
+ }
neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/granite/hf-internal-testing/tiny-random-GraniteForCausalLM/60271363e0e147450b0f.json ADDED
@@ -0,0 +1,73 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "hf-internal-testing/tiny-random-GraniteForCausalLM",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "GraniteForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "attention_multiplier": 1.0,
11
+ "embedding_multiplier": 1.0,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 32,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 64,
16
+ "logits_scaling": 1.0,
17
+ "max_position_embeddings": 2048,
18
+ "mlp_bias": false,
19
+ "model_type": "granite",
20
+ "neuron": {
21
+ "_serialized_key": "NxDNeuronConfig",
22
+ "async_mode": false,
23
+ "attn_kernel_enabled": false,
24
+ "batch_size": 1,
25
+ "capacity_factor": null,
26
+ "cc_pipeline_tiling_factor": 2,
27
+ "checkpoint_id": "hf-internal-testing/tiny-random-GraniteForCausalLM",
28
+ "checkpoint_revision": "c3074ebc0ac2fe545305f5e5f6cce2cc9b2aa0c5",
29
+ "continuous_batching": false,
30
+ "enable_bucketing": false,
31
+ "ep_degree": 1,
32
+ "flash_decoding_enabled": false,
33
+ "fused_qkv": true,
34
+ "glu_mlp": true,
35
+ "is_chunked_prefill": false,
36
+ "local_ranks_size": 2,
37
+ "logical_nc_config": 1,
38
+ "max_batch_size": 1,
39
+ "max_context_length": 100,
40
+ "max_topk": 256,
41
+ "mlp_kernel_enabled": false,
42
+ "mlp_kernel_fuse_residual_add": false,
43
+ "n_active_tokens": 100,
44
+ "neuronxcc_version": "2.19.8089.0+8ab9f450",
45
+ "num_cores_per_group": 1,
46
+ "on_device_sampling": true,
47
+ "optimum_neuron_version": "0.3.1.dev1",
48
+ "output_logits": false,
49
+ "padding_side": "right",
50
+ "pp_degree": 1,
51
+ "qk_layernorm": false,
52
+ "qkv_kernel_enabled": false,
53
+ "rpl_reduce_dtype": "float16",
54
+ "sequence_length": 100,
55
+ "sequence_parallel_enabled": false,
56
+ "speculation_length": 0,
57
+ "start_rank_id": 0,
58
+ "target": null,
59
+ "torch_dtype": "float16",
60
+ "tp_degree": 2,
61
+ "vocab_parallel": false
62
+ },
63
+ "num_attention_heads": 4,
64
+ "num_hidden_layers": 2,
65
+ "num_key_value_heads": 4,
66
+ "residual_multiplier": 1.0,
67
+ "rms_norm_eps": 1e-06,
68
+ "rope_scaling": null,
69
+ "rope_theta": 10000.0,
70
+ "tie_word_embeddings": false,
71
+ "use_cache": true,
72
+ "vocab_size": 49152
73
+ }
neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/granite/hf-internal-testing/tiny-random-GraniteForCausalLM/a9e5f04b708cd397fbdd.json ADDED
@@ -0,0 +1,73 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "hf-internal-testing/tiny-random-GraniteForCausalLM",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "GraniteForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "attention_multiplier": 1.0,
11
+ "embedding_multiplier": 1.0,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 32,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 64,
16
+ "logits_scaling": 1.0,
17
+ "max_position_embeddings": 2048,
18
+ "mlp_bias": false,
19
+ "model_type": "granite",
20
+ "neuron": {
21
+ "_serialized_key": "NxDNeuronConfig",
22
+ "async_mode": false,
23
+ "attn_kernel_enabled": false,
24
+ "batch_size": 2,
25
+ "capacity_factor": null,
26
+ "cc_pipeline_tiling_factor": 2,
27
+ "checkpoint_id": "hf-internal-testing/tiny-random-GraniteForCausalLM",
28
+ "checkpoint_revision": "c3074ebc0ac2fe545305f5e5f6cce2cc9b2aa0c5",
29
+ "continuous_batching": true,
30
+ "enable_bucketing": false,
31
+ "ep_degree": 1,
32
+ "flash_decoding_enabled": false,
33
+ "fused_qkv": true,
34
+ "glu_mlp": true,
35
+ "is_chunked_prefill": false,
36
+ "local_ranks_size": 2,
37
+ "logical_nc_config": 1,
38
+ "max_batch_size": 2,
39
+ "max_context_length": 100,
40
+ "max_topk": 256,
41
+ "mlp_kernel_enabled": false,
42
+ "mlp_kernel_fuse_residual_add": false,
43
+ "n_active_tokens": 100,
44
+ "neuronxcc_version": "2.19.8089.0+8ab9f450",
45
+ "num_cores_per_group": 1,
46
+ "on_device_sampling": true,
47
+ "optimum_neuron_version": "0.3.1.dev1",
48
+ "output_logits": false,
49
+ "padding_side": "right",
50
+ "pp_degree": 1,
51
+ "qk_layernorm": false,
52
+ "qkv_kernel_enabled": false,
53
+ "rpl_reduce_dtype": "float16",
54
+ "sequence_length": 100,
55
+ "sequence_parallel_enabled": false,
56
+ "speculation_length": 0,
57
+ "start_rank_id": 0,
58
+ "target": null,
59
+ "torch_dtype": "float16",
60
+ "tp_degree": 2,
61
+ "vocab_parallel": false
62
+ },
63
+ "num_attention_heads": 4,
64
+ "num_hidden_layers": 2,
65
+ "num_key_value_heads": 4,
66
+ "residual_multiplier": 1.0,
67
+ "rms_norm_eps": 1e-06,
68
+ "rope_scaling": null,
69
+ "rope_theta": 10000.0,
70
+ "tie_word_embeddings": false,
71
+ "use_cache": true,
72
+ "vocab_size": 49152
73
+ }
neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/granite/hf-internal-testing/tiny-random-GraniteForCausalLM/de79a3968941a6d3864f.json ADDED
@@ -0,0 +1,73 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "hf-internal-testing/tiny-random-GraniteForCausalLM",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "GraniteForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "attention_multiplier": 1.0,
11
+ "embedding_multiplier": 1.0,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 32,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 64,
16
+ "logits_scaling": 1.0,
17
+ "max_position_embeddings": 2048,
18
+ "mlp_bias": false,
19
+ "model_type": "granite",
20
+ "neuron": {
21
+ "_serialized_key": "NxDNeuronConfig",
22
+ "async_mode": false,
23
+ "attn_kernel_enabled": false,
24
+ "batch_size": 1,
25
+ "capacity_factor": null,
26
+ "cc_pipeline_tiling_factor": 2,
27
+ "checkpoint_id": "hf-internal-testing/tiny-random-GraniteForCausalLM",
28
+ "checkpoint_revision": "c3074ebc0ac2fe545305f5e5f6cce2cc9b2aa0c5",
29
+ "continuous_batching": false,
30
+ "enable_bucketing": false,
31
+ "ep_degree": 1,
32
+ "flash_decoding_enabled": false,
33
+ "fused_qkv": true,
34
+ "glu_mlp": true,
35
+ "is_chunked_prefill": false,
36
+ "local_ranks_size": 2,
37
+ "logical_nc_config": 1,
38
+ "max_batch_size": 1,
39
+ "max_context_length": 100,
40
+ "max_topk": 256,
41
+ "mlp_kernel_enabled": false,
42
+ "mlp_kernel_fuse_residual_add": false,
43
+ "n_active_tokens": 100,
44
+ "neuronxcc_version": "2.19.8089.0+8ab9f450",
45
+ "num_cores_per_group": 1,
46
+ "on_device_sampling": true,
47
+ "optimum_neuron_version": "0.3.1.dev1",
48
+ "output_logits": false,
49
+ "padding_side": "right",
50
+ "pp_degree": 1,
51
+ "qk_layernorm": false,
52
+ "qkv_kernel_enabled": false,
53
+ "rpl_reduce_dtype": "bfloat16",
54
+ "sequence_length": 100,
55
+ "sequence_parallel_enabled": false,
56
+ "speculation_length": 0,
57
+ "start_rank_id": 0,
58
+ "target": null,
59
+ "torch_dtype": "bfloat16",
60
+ "tp_degree": 2,
61
+ "vocab_parallel": false
62
+ },
63
+ "num_attention_heads": 4,
64
+ "num_hidden_layers": 2,
65
+ "num_key_value_heads": 4,
66
+ "residual_multiplier": 1.0,
67
+ "rms_norm_eps": 1e-06,
68
+ "rope_scaling": null,
69
+ "rope_theta": 10000.0,
70
+ "tie_word_embeddings": false,
71
+ "use_cache": true,
72
+ "vocab_size": 49152
73
+ }
neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/llama/llamafactory/tiny-random-Llama-3/28b90633a77c30e8a187.json ADDED
@@ -0,0 +1,77 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "llamafactory/tiny-random-Llama-3",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "head_dim": 4,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 16,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 64,
15
+ "max_position_embeddings": 131072,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "neuron": {
19
+ "_serialized_key": "NxDNeuronConfig",
20
+ "async_mode": false,
21
+ "attn_kernel_enabled": false,
22
+ "batch_size": 1,
23
+ "capacity_factor": null,
24
+ "cc_pipeline_tiling_factor": 2,
25
+ "checkpoint_id": "llamafactory/tiny-random-Llama-3",
26
+ "checkpoint_revision": "bf2a2e3bf199ad2ee96f02a3c00246c608db22a8",
27
+ "continuous_batching": false,
28
+ "enable_bucketing": false,
29
+ "ep_degree": 1,
30
+ "flash_decoding_enabled": false,
31
+ "fused_qkv": true,
32
+ "glu_mlp": true,
33
+ "is_chunked_prefill": false,
34
+ "local_ranks_size": 2,
35
+ "logical_nc_config": 1,
36
+ "max_batch_size": 1,
37
+ "max_context_length": 100,
38
+ "max_topk": 256,
39
+ "mlp_kernel_enabled": false,
40
+ "mlp_kernel_fuse_residual_add": false,
41
+ "n_active_tokens": 100,
42
+ "neuronxcc_version": "2.19.8089.0+8ab9f450",
43
+ "num_cores_per_group": 1,
44
+ "on_device_sampling": true,
45
+ "optimum_neuron_version": "0.3.1.dev1",
46
+ "output_logits": false,
47
+ "padding_side": "right",
48
+ "pp_degree": 1,
49
+ "qk_layernorm": false,
50
+ "qkv_kernel_enabled": false,
51
+ "rpl_reduce_dtype": "float16",
52
+ "sequence_length": 100,
53
+ "sequence_parallel_enabled": false,
54
+ "speculation_length": 0,
55
+ "start_rank_id": 0,
56
+ "target": null,
57
+ "torch_dtype": "float16",
58
+ "tp_degree": 2,
59
+ "vocab_parallel": false
60
+ },
61
+ "num_attention_heads": 4,
62
+ "num_hidden_layers": 2,
63
+ "num_key_value_heads": 4,
64
+ "pretraining_tp": 1,
65
+ "rms_norm_eps": 1e-05,
66
+ "rope_scaling": {
67
+ "factor": 8.0,
68
+ "high_freq_factor": 4.0,
69
+ "low_freq_factor": 1.0,
70
+ "original_max_position_embeddings": 8192,
71
+ "rope_type": "llama3"
72
+ },
73
+ "rope_theta": 500000.0,
74
+ "tie_word_embeddings": false,
75
+ "use_cache": true,
76
+ "vocab_size": 128256
77
+ }
neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/llama/llamafactory/tiny-random-Llama-3/88a97541d018d7aa0b29.json ADDED
@@ -0,0 +1,77 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "llamafactory/tiny-random-Llama-3",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "head_dim": 4,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 16,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 64,
15
+ "max_position_embeddings": 131072,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "neuron": {
19
+ "_serialized_key": "NxDNeuronConfig",
20
+ "async_mode": false,
21
+ "attn_kernel_enabled": false,
22
+ "batch_size": 1,
23
+ "capacity_factor": null,
24
+ "cc_pipeline_tiling_factor": 2,
25
+ "checkpoint_id": "llamafactory/tiny-random-Llama-3",
26
+ "checkpoint_revision": "bf2a2e3bf199ad2ee96f02a3c00246c608db22a8",
27
+ "continuous_batching": false,
28
+ "enable_bucketing": false,
29
+ "ep_degree": 1,
30
+ "flash_decoding_enabled": false,
31
+ "fused_qkv": true,
32
+ "glu_mlp": true,
33
+ "is_chunked_prefill": false,
34
+ "local_ranks_size": 2,
35
+ "logical_nc_config": 1,
36
+ "max_batch_size": 1,
37
+ "max_context_length": 100,
38
+ "max_topk": 256,
39
+ "mlp_kernel_enabled": false,
40
+ "mlp_kernel_fuse_residual_add": false,
41
+ "n_active_tokens": 100,
42
+ "neuronxcc_version": "2.19.8089.0+8ab9f450",
43
+ "num_cores_per_group": 1,
44
+ "on_device_sampling": true,
45
+ "optimum_neuron_version": "0.3.1.dev1",
46
+ "output_logits": false,
47
+ "padding_side": "right",
48
+ "pp_degree": 1,
49
+ "qk_layernorm": false,
50
+ "qkv_kernel_enabled": false,
51
+ "rpl_reduce_dtype": "bfloat16",
52
+ "sequence_length": 100,
53
+ "sequence_parallel_enabled": false,
54
+ "speculation_length": 0,
55
+ "start_rank_id": 0,
56
+ "target": null,
57
+ "torch_dtype": "bfloat16",
58
+ "tp_degree": 2,
59
+ "vocab_parallel": false
60
+ },
61
+ "num_attention_heads": 4,
62
+ "num_hidden_layers": 2,
63
+ "num_key_value_heads": 4,
64
+ "pretraining_tp": 1,
65
+ "rms_norm_eps": 1e-05,
66
+ "rope_scaling": {
67
+ "factor": 8.0,
68
+ "high_freq_factor": 4.0,
69
+ "low_freq_factor": 1.0,
70
+ "original_max_position_embeddings": 8192,
71
+ "rope_type": "llama3"
72
+ },
73
+ "rope_theta": 500000.0,
74
+ "tie_word_embeddings": false,
75
+ "use_cache": true,
76
+ "vocab_size": 128256
77
+ }
neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/llama/llamafactory/tiny-random-Llama-3/d8f338a72f4bb0a4682d.json ADDED
@@ -0,0 +1,77 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "llamafactory/tiny-random-Llama-3",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "head_dim": 4,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 16,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 64,
15
+ "max_position_embeddings": 131072,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "neuron": {
19
+ "_serialized_key": "NxDNeuronConfig",
20
+ "async_mode": false,
21
+ "attn_kernel_enabled": false,
22
+ "batch_size": 2,
23
+ "capacity_factor": null,
24
+ "cc_pipeline_tiling_factor": 2,
25
+ "checkpoint_id": "llamafactory/tiny-random-Llama-3",
26
+ "checkpoint_revision": "bf2a2e3bf199ad2ee96f02a3c00246c608db22a8",
27
+ "continuous_batching": true,
28
+ "enable_bucketing": false,
29
+ "ep_degree": 1,
30
+ "flash_decoding_enabled": false,
31
+ "fused_qkv": true,
32
+ "glu_mlp": true,
33
+ "is_chunked_prefill": false,
34
+ "local_ranks_size": 2,
35
+ "logical_nc_config": 1,
36
+ "max_batch_size": 2,
37
+ "max_context_length": 100,
38
+ "max_topk": 256,
39
+ "mlp_kernel_enabled": false,
40
+ "mlp_kernel_fuse_residual_add": false,
41
+ "n_active_tokens": 100,
42
+ "neuronxcc_version": "2.19.8089.0+8ab9f450",
43
+ "num_cores_per_group": 1,
44
+ "on_device_sampling": true,
45
+ "optimum_neuron_version": "0.3.1.dev1",
46
+ "output_logits": false,
47
+ "padding_side": "right",
48
+ "pp_degree": 1,
49
+ "qk_layernorm": false,
50
+ "qkv_kernel_enabled": false,
51
+ "rpl_reduce_dtype": "float16",
52
+ "sequence_length": 100,
53
+ "sequence_parallel_enabled": false,
54
+ "speculation_length": 0,
55
+ "start_rank_id": 0,
56
+ "target": null,
57
+ "torch_dtype": "float16",
58
+ "tp_degree": 2,
59
+ "vocab_parallel": false
60
+ },
61
+ "num_attention_heads": 4,
62
+ "num_hidden_layers": 2,
63
+ "num_key_value_heads": 4,
64
+ "pretraining_tp": 1,
65
+ "rms_norm_eps": 1e-05,
66
+ "rope_scaling": {
67
+ "factor": 8.0,
68
+ "high_freq_factor": 4.0,
69
+ "low_freq_factor": 1.0,
70
+ "original_max_position_embeddings": 8192,
71
+ "rope_type": "llama3"
72
+ },
73
+ "rope_theta": 500000.0,
74
+ "tie_word_embeddings": false,
75
+ "use_cache": true,
76
+ "vocab_size": 128256
77
+ }
neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/llama/unsloth/Llama-3.2-1B-Instruct/f7ed18f9ae29fab731ad.json ADDED
@@ -0,0 +1,78 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "head_dim": 64,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 2048,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 8192,
15
+ "max_position_embeddings": 131072,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "neuron": {
19
+ "_serialized_key": "NxDNeuronConfig",
20
+ "async_mode": false,
21
+ "attn_kernel_enabled": false,
22
+ "batch_size": 4,
23
+ "capacity_factor": null,
24
+ "cc_pipeline_tiling_factor": 2,
25
+ "checkpoint_id": "unsloth/Llama-3.2-1B-Instruct",
26
+ "checkpoint_revision": "5a8abab4a5d6f164389b1079fb721cfab8d7126c",
27
+ "continuous_batching": true,
28
+ "enable_bucketing": false,
29
+ "ep_degree": 1,
30
+ "flash_decoding_enabled": false,
31
+ "fused_qkv": true,
32
+ "glu_mlp": true,
33
+ "is_chunked_prefill": false,
34
+ "local_ranks_size": 2,
35
+ "logical_nc_config": 1,
36
+ "max_batch_size": 4,
37
+ "max_context_length": 4096,
38
+ "max_topk": 256,
39
+ "mlp_kernel_enabled": false,
40
+ "mlp_kernel_fuse_residual_add": false,
41
+ "n_active_tokens": 4096,
42
+ "neuronxcc_version": "2.19.8089.0+8ab9f450",
43
+ "num_cores_per_group": 1,
44
+ "on_device_sampling": true,
45
+ "optimum_neuron_version": "0.3.1.dev1",
46
+ "output_logits": false,
47
+ "padding_side": "right",
48
+ "pp_degree": 1,
49
+ "qk_layernorm": false,
50
+ "qkv_kernel_enabled": false,
51
+ "rpl_reduce_dtype": "float16",
52
+ "sequence_length": 4096,
53
+ "sequence_parallel_enabled": false,
54
+ "speculation_length": 0,
55
+ "start_rank_id": 0,
56
+ "target": null,
57
+ "torch_dtype": "float16",
58
+ "tp_degree": 2,
59
+ "vocab_parallel": false
60
+ },
61
+ "num_attention_heads": 32,
62
+ "num_hidden_layers": 16,
63
+ "num_key_value_heads": 8,
64
+ "pretraining_tp": 1,
65
+ "rms_norm_eps": 1e-05,
66
+ "rope_scaling": {
67
+ "factor": 32.0,
68
+ "high_freq_factor": 4.0,
69
+ "low_freq_factor": 1.0,
70
+ "original_max_position_embeddings": 8192,
71
+ "rope_type": "llama3"
72
+ },
73
+ "rope_theta": 500000.0,
74
+ "tie_word_embeddings": true,
75
+ "unsloth_fixed": true,
76
+ "use_cache": true,
77
+ "vocab_size": 128256
78
+ }
neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/phi3/yujiepan/phi-4-tiny-random/0d648a236d7cc578c5a8.json ADDED
@@ -0,0 +1,74 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "yujiepan/phi-4-tiny-random",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Phi3ForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "auto_map": {},
11
+ "embd_pdrop": 0.0,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 16,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 32,
16
+ "max_position_embeddings": 16384,
17
+ "model_type": "phi3",
18
+ "neuron": {
19
+ "_serialized_key": "NxDNeuronConfig",
20
+ "async_mode": false,
21
+ "attn_kernel_enabled": false,
22
+ "batch_size": 1,
23
+ "capacity_factor": null,
24
+ "cc_pipeline_tiling_factor": 2,
25
+ "checkpoint_id": "yujiepan/phi-4-tiny-random",
26
+ "checkpoint_revision": "18a9a1168dc97ac6d128f811925670c275610f5a",
27
+ "continuous_batching": false,
28
+ "enable_bucketing": false,
29
+ "ep_degree": 1,
30
+ "flash_decoding_enabled": false,
31
+ "fused_qkv": true,
32
+ "glu_mlp": true,
33
+ "is_chunked_prefill": false,
34
+ "local_ranks_size": 2,
35
+ "logical_nc_config": 1,
36
+ "max_batch_size": 1,
37
+ "max_context_length": 100,
38
+ "max_topk": 256,
39
+ "mlp_kernel_enabled": false,
40
+ "mlp_kernel_fuse_residual_add": false,
41
+ "n_active_tokens": 100,
42
+ "neuronxcc_version": "2.19.8089.0+8ab9f450",
43
+ "num_cores_per_group": 1,
44
+ "on_device_sampling": true,
45
+ "optimum_neuron_version": "0.3.1.dev1",
46
+ "output_logits": false,
47
+ "padding_side": "right",
48
+ "pp_degree": 1,
49
+ "qk_layernorm": false,
50
+ "qkv_kernel_enabled": false,
51
+ "rpl_reduce_dtype": "bfloat16",
52
+ "sequence_length": 100,
53
+ "sequence_parallel_enabled": false,
54
+ "speculation_length": 0,
55
+ "start_rank_id": 0,
56
+ "target": null,
57
+ "torch_dtype": "bfloat16",
58
+ "tp_degree": 2,
59
+ "vocab_parallel": false
60
+ },
61
+ "num_attention_heads": 2,
62
+ "num_hidden_layers": 2,
63
+ "num_key_value_heads": 1,
64
+ "original_max_position_embeddings": 16384,
65
+ "partial_rotary_factor": 1.0,
66
+ "resid_pdrop": 0.0,
67
+ "rms_norm_eps": 1e-05,
68
+ "rope_scaling": null,
69
+ "rope_theta": 250000,
70
+ "sliding_window": null,
71
+ "tie_word_embeddings": false,
72
+ "use_cache": true,
73
+ "vocab_size": 100352
74
+ }
neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/phi3/yujiepan/phi-4-tiny-random/1228009b3909ec8d9751.json ADDED
@@ -0,0 +1,74 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "yujiepan/phi-4-tiny-random",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Phi3ForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "auto_map": {},
11
+ "embd_pdrop": 0.0,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 16,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 32,
16
+ "max_position_embeddings": 16384,
17
+ "model_type": "phi3",
18
+ "neuron": {
19
+ "_serialized_key": "NxDNeuronConfig",
20
+ "async_mode": false,
21
+ "attn_kernel_enabled": false,
22
+ "batch_size": 2,
23
+ "capacity_factor": null,
24
+ "cc_pipeline_tiling_factor": 2,
25
+ "checkpoint_id": "yujiepan/phi-4-tiny-random",
26
+ "checkpoint_revision": "18a9a1168dc97ac6d128f811925670c275610f5a",
27
+ "continuous_batching": true,
28
+ "enable_bucketing": false,
29
+ "ep_degree": 1,
30
+ "flash_decoding_enabled": false,
31
+ "fused_qkv": true,
32
+ "glu_mlp": true,
33
+ "is_chunked_prefill": false,
34
+ "local_ranks_size": 2,
35
+ "logical_nc_config": 1,
36
+ "max_batch_size": 2,
37
+ "max_context_length": 100,
38
+ "max_topk": 256,
39
+ "mlp_kernel_enabled": false,
40
+ "mlp_kernel_fuse_residual_add": false,
41
+ "n_active_tokens": 100,
42
+ "neuronxcc_version": "2.19.8089.0+8ab9f450",
43
+ "num_cores_per_group": 1,
44
+ "on_device_sampling": true,
45
+ "optimum_neuron_version": "0.3.1.dev1",
46
+ "output_logits": false,
47
+ "padding_side": "right",
48
+ "pp_degree": 1,
49
+ "qk_layernorm": false,
50
+ "qkv_kernel_enabled": false,
51
+ "rpl_reduce_dtype": "float16",
52
+ "sequence_length": 100,
53
+ "sequence_parallel_enabled": false,
54
+ "speculation_length": 0,
55
+ "start_rank_id": 0,
56
+ "target": null,
57
+ "torch_dtype": "float16",
58
+ "tp_degree": 2,
59
+ "vocab_parallel": false
60
+ },
61
+ "num_attention_heads": 2,
62
+ "num_hidden_layers": 2,
63
+ "num_key_value_heads": 1,
64
+ "original_max_position_embeddings": 16384,
65
+ "partial_rotary_factor": 1.0,
66
+ "resid_pdrop": 0.0,
67
+ "rms_norm_eps": 1e-05,
68
+ "rope_scaling": null,
69
+ "rope_theta": 250000,
70
+ "sliding_window": null,
71
+ "tie_word_embeddings": false,
72
+ "use_cache": true,
73
+ "vocab_size": 100352
74
+ }
neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/phi3/yujiepan/phi-4-tiny-random/1c5104d89e025d3fb8db.json ADDED
@@ -0,0 +1,74 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "yujiepan/phi-4-tiny-random",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Phi3ForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "auto_map": {},
11
+ "embd_pdrop": 0.0,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 16,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 32,
16
+ "max_position_embeddings": 16384,
17
+ "model_type": "phi3",
18
+ "neuron": {
19
+ "_serialized_key": "NxDNeuronConfig",
20
+ "async_mode": false,
21
+ "attn_kernel_enabled": false,
22
+ "batch_size": 1,
23
+ "capacity_factor": null,
24
+ "cc_pipeline_tiling_factor": 2,
25
+ "checkpoint_id": "yujiepan/phi-4-tiny-random",
26
+ "checkpoint_revision": "18a9a1168dc97ac6d128f811925670c275610f5a",
27
+ "continuous_batching": false,
28
+ "enable_bucketing": false,
29
+ "ep_degree": 1,
30
+ "flash_decoding_enabled": false,
31
+ "fused_qkv": true,
32
+ "glu_mlp": true,
33
+ "is_chunked_prefill": false,
34
+ "local_ranks_size": 2,
35
+ "logical_nc_config": 1,
36
+ "max_batch_size": 1,
37
+ "max_context_length": 100,
38
+ "max_topk": 256,
39
+ "mlp_kernel_enabled": false,
40
+ "mlp_kernel_fuse_residual_add": false,
41
+ "n_active_tokens": 100,
42
+ "neuronxcc_version": "2.19.8089.0+8ab9f450",
43
+ "num_cores_per_group": 1,
44
+ "on_device_sampling": true,
45
+ "optimum_neuron_version": "0.3.1.dev1",
46
+ "output_logits": false,
47
+ "padding_side": "right",
48
+ "pp_degree": 1,
49
+ "qk_layernorm": false,
50
+ "qkv_kernel_enabled": false,
51
+ "rpl_reduce_dtype": "float16",
52
+ "sequence_length": 100,
53
+ "sequence_parallel_enabled": false,
54
+ "speculation_length": 0,
55
+ "start_rank_id": 0,
56
+ "target": null,
57
+ "torch_dtype": "float16",
58
+ "tp_degree": 2,
59
+ "vocab_parallel": false
60
+ },
61
+ "num_attention_heads": 2,
62
+ "num_hidden_layers": 2,
63
+ "num_key_value_heads": 1,
64
+ "original_max_position_embeddings": 16384,
65
+ "partial_rotary_factor": 1.0,
66
+ "resid_pdrop": 0.0,
67
+ "rms_norm_eps": 1e-05,
68
+ "rope_scaling": null,
69
+ "rope_theta": 250000,
70
+ "sliding_window": null,
71
+ "tie_word_embeddings": false,
72
+ "use_cache": true,
73
+ "vocab_size": 100352
74
+ }
neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/qwen2/yujiepan/qwen2.5-128k-tiny-random/4f65451820dd0eadcdca.json ADDED
@@ -0,0 +1,79 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "yujiepan/qwen2.5-128k-tiny-random",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Qwen2ForCausalLM"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "hidden_act": "silu",
10
+ "hidden_size": 8,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 16,
13
+ "layer_types": [
14
+ "full_attention",
15
+ "full_attention"
16
+ ],
17
+ "max_position_embeddings": 32768,
18
+ "max_window_layers": 1,
19
+ "model_type": "qwen2",
20
+ "neuron": {
21
+ "_serialized_key": "NxDNeuronConfig",
22
+ "async_mode": false,
23
+ "attn_kernel_enabled": false,
24
+ "batch_size": 2,
25
+ "capacity_factor": null,
26
+ "cc_pipeline_tiling_factor": 2,
27
+ "checkpoint_id": "yujiepan/qwen2.5-128k-tiny-random",
28
+ "checkpoint_revision": "c8296d4ca3f87782876d2382fbb6481d1beb8ef0",
29
+ "continuous_batching": true,
30
+ "enable_bucketing": false,
31
+ "ep_degree": 1,
32
+ "flash_decoding_enabled": false,
33
+ "fused_qkv": false,
34
+ "glu_mlp": true,
35
+ "is_chunked_prefill": false,
36
+ "local_ranks_size": 2,
37
+ "logical_nc_config": 1,
38
+ "max_batch_size": 2,
39
+ "max_context_length": 100,
40
+ "max_topk": 256,
41
+ "mlp_kernel_enabled": false,
42
+ "mlp_kernel_fuse_residual_add": false,
43
+ "n_active_tokens": 100,
44
+ "neuronxcc_version": "2.19.8089.0+8ab9f450",
45
+ "num_cores_per_group": 1,
46
+ "on_device_sampling": false,
47
+ "optimum_neuron_version": "0.3.1.dev1",
48
+ "output_logits": false,
49
+ "padding_side": "right",
50
+ "pp_degree": 1,
51
+ "qk_layernorm": false,
52
+ "qkv_kernel_enabled": false,
53
+ "rpl_reduce_dtype": "float16",
54
+ "sequence_length": 100,
55
+ "sequence_parallel_enabled": false,
56
+ "speculation_length": 0,
57
+ "start_rank_id": 0,
58
+ "target": null,
59
+ "torch_dtype": "float16",
60
+ "tp_degree": 2,
61
+ "vocab_parallel": false
62
+ },
63
+ "num_attention_heads": 4,
64
+ "num_hidden_layers": 2,
65
+ "num_key_value_heads": 2,
66
+ "rms_norm_eps": 1e-06,
67
+ "rope_scaling": {
68
+ "factor": 4.0,
69
+ "original_max_position_embeddings": 32768,
70
+ "rope_type": "yarn",
71
+ "type": "yarn"
72
+ },
73
+ "rope_theta": 1000000.0,
74
+ "sliding_window": null,
75
+ "tie_word_embeddings": false,
76
+ "use_cache": true,
77
+ "use_sliding_window": false,
78
+ "vocab_size": 152064
79
+ }
neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/qwen2/yujiepan/qwen2.5-128k-tiny-random/d7566c1b656e75712c0c.json ADDED
@@ -0,0 +1,79 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "yujiepan/qwen2.5-128k-tiny-random",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Qwen2ForCausalLM"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "hidden_act": "silu",
10
+ "hidden_size": 8,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 16,
13
+ "layer_types": [
14
+ "full_attention",
15
+ "full_attention"
16
+ ],
17
+ "max_position_embeddings": 32768,
18
+ "max_window_layers": 1,
19
+ "model_type": "qwen2",
20
+ "neuron": {
21
+ "_serialized_key": "NxDNeuronConfig",
22
+ "async_mode": false,
23
+ "attn_kernel_enabled": false,
24
+ "batch_size": 1,
25
+ "capacity_factor": null,
26
+ "cc_pipeline_tiling_factor": 2,
27
+ "checkpoint_id": "yujiepan/qwen2.5-128k-tiny-random",
28
+ "checkpoint_revision": "c8296d4ca3f87782876d2382fbb6481d1beb8ef0",
29
+ "continuous_batching": false,
30
+ "enable_bucketing": false,
31
+ "ep_degree": 1,
32
+ "flash_decoding_enabled": false,
33
+ "fused_qkv": false,
34
+ "glu_mlp": true,
35
+ "is_chunked_prefill": false,
36
+ "local_ranks_size": 2,
37
+ "logical_nc_config": 1,
38
+ "max_batch_size": 1,
39
+ "max_context_length": 100,
40
+ "max_topk": 256,
41
+ "mlp_kernel_enabled": false,
42
+ "mlp_kernel_fuse_residual_add": false,
43
+ "n_active_tokens": 100,
44
+ "neuronxcc_version": "2.19.8089.0+8ab9f450",
45
+ "num_cores_per_group": 1,
46
+ "on_device_sampling": true,
47
+ "optimum_neuron_version": "0.3.1.dev1",
48
+ "output_logits": false,
49
+ "padding_side": "right",
50
+ "pp_degree": 1,
51
+ "qk_layernorm": false,
52
+ "qkv_kernel_enabled": false,
53
+ "rpl_reduce_dtype": "bfloat16",
54
+ "sequence_length": 100,
55
+ "sequence_parallel_enabled": false,
56
+ "speculation_length": 0,
57
+ "start_rank_id": 0,
58
+ "target": null,
59
+ "torch_dtype": "bfloat16",
60
+ "tp_degree": 2,
61
+ "vocab_parallel": false
62
+ },
63
+ "num_attention_heads": 4,
64
+ "num_hidden_layers": 2,
65
+ "num_key_value_heads": 2,
66
+ "rms_norm_eps": 1e-06,
67
+ "rope_scaling": {
68
+ "factor": 4.0,
69
+ "original_max_position_embeddings": 32768,
70
+ "rope_type": "yarn",
71
+ "type": "yarn"
72
+ },
73
+ "rope_theta": 1000000.0,
74
+ "sliding_window": null,
75
+ "tie_word_embeddings": false,
76
+ "use_cache": true,
77
+ "use_sliding_window": false,
78
+ "vocab_size": 152064
79
+ }
neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev1/qwen2/yujiepan/qwen2.5-128k-tiny-random/f47233e40681eaec5807.json ADDED
@@ -0,0 +1,79 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "yujiepan/qwen2.5-128k-tiny-random",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Qwen2ForCausalLM"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "hidden_act": "silu",
10
+ "hidden_size": 8,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 16,
13
+ "layer_types": [
14
+ "full_attention",
15
+ "full_attention"
16
+ ],
17
+ "max_position_embeddings": 32768,
18
+ "max_window_layers": 1,
19
+ "model_type": "qwen2",
20
+ "neuron": {
21
+ "_serialized_key": "NxDNeuronConfig",
22
+ "async_mode": false,
23
+ "attn_kernel_enabled": false,
24
+ "batch_size": 1,
25
+ "capacity_factor": null,
26
+ "cc_pipeline_tiling_factor": 2,
27
+ "checkpoint_id": "yujiepan/qwen2.5-128k-tiny-random",
28
+ "checkpoint_revision": "c8296d4ca3f87782876d2382fbb6481d1beb8ef0",
29
+ "continuous_batching": false,
30
+ "enable_bucketing": false,
31
+ "ep_degree": 1,
32
+ "flash_decoding_enabled": false,
33
+ "fused_qkv": false,
34
+ "glu_mlp": true,
35
+ "is_chunked_prefill": false,
36
+ "local_ranks_size": 2,
37
+ "logical_nc_config": 1,
38
+ "max_batch_size": 1,
39
+ "max_context_length": 100,
40
+ "max_topk": 256,
41
+ "mlp_kernel_enabled": false,
42
+ "mlp_kernel_fuse_residual_add": false,
43
+ "n_active_tokens": 100,
44
+ "neuronxcc_version": "2.19.8089.0+8ab9f450",
45
+ "num_cores_per_group": 1,
46
+ "on_device_sampling": true,
47
+ "optimum_neuron_version": "0.3.1.dev1",
48
+ "output_logits": false,
49
+ "padding_side": "right",
50
+ "pp_degree": 1,
51
+ "qk_layernorm": false,
52
+ "qkv_kernel_enabled": false,
53
+ "rpl_reduce_dtype": "float16",
54
+ "sequence_length": 100,
55
+ "sequence_parallel_enabled": false,
56
+ "speculation_length": 0,
57
+ "start_rank_id": 0,
58
+ "target": null,
59
+ "torch_dtype": "float16",
60
+ "tp_degree": 2,
61
+ "vocab_parallel": false
62
+ },
63
+ "num_attention_heads": 4,
64
+ "num_hidden_layers": 2,
65
+ "num_key_value_heads": 2,
66
+ "rms_norm_eps": 1e-06,
67
+ "rope_scaling": {
68
+ "factor": 4.0,
69
+ "original_max_position_embeddings": 32768,
70
+ "rope_type": "yarn",
71
+ "type": "yarn"
72
+ },
73
+ "rope_theta": 1000000.0,
74
+ "sliding_window": null,
75
+ "tie_word_embeddings": false,
76
+ "use_cache": true,
77
+ "use_sliding_window": false,
78
+ "vocab_size": 152064
79
+ }
neuronxcc-2.19.8089.0+8ab9f450/MODULE_018530e1260c6c7cee60+ed72d204/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_model/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.19.8089.0+8ab9f450/MODULE_018530e1260c6c7cee60+ed72d204/model.done ADDED
File without changes
neuronxcc-2.19.8089.0+8ab9f450/MODULE_018530e1260c6c7cee60+ed72d204/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:26733925702af59a7abc8f84c20ccf865775eabb94345ce202abfad875a9ad01
3
+ size 888165
neuronxcc-2.19.8089.0+8ab9f450/MODULE_018530e1260c6c7cee60+ed72d204/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4621fb761c7e4fe70f3db98fcf513d48a57cd3df25ce6cabf49c41e0cd4408f4
3
+ size 14828544
neuronxcc-2.19.8089.0+8ab9f450/MODULE_16dadb9b6ab4bb044c1e+a9d440f5/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_model/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
neuronxcc-2.19.8089.0+8ab9f450/MODULE_16dadb9b6ab4bb044c1e+a9d440f5/model.done ADDED
File without changes
neuronxcc-2.19.8089.0+8ab9f450/MODULE_16dadb9b6ab4bb044c1e+a9d440f5/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:72287a987834149a2335abf62583736027e879d554dc31cde709dfa136503eee
3
+ size 803944
neuronxcc-2.19.8089.0+8ab9f450/MODULE_16dadb9b6ab4bb044c1e+a9d440f5/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c028818a0dff638ebbd1d62fc8de93d62b5dce39299d501c1c94bf17fec57c6f
3
+ size 3830784
neuronxcc-2.19.8089.0+8ab9f450/MODULE_16dadb9b6ab4bb044c1e+a9d440f5/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:45b8f7aa6c0e42823cb6937ca044c0857914f7915bc5da99db9191908b277872
3
+ size 4025809
neuronxcc-2.19.8089.0+8ab9f450/MODULE_1d2b7dee00e90e52700e+431f5505/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ "--model-type=transformer -O1 --lnc=1 --internal-hlo2tensorizer-options=--experimental-unsafe-fp8e4m3fn-as-fp8e4m3 --logfile=/tmp/nxd_model/layout_opt/log-neuron-cc.txt"
neuronxcc-2.19.8089.0+8ab9f450/MODULE_1d2b7dee00e90e52700e+431f5505/model.done ADDED
File without changes
neuronxcc-2.19.8089.0+8ab9f450/MODULE_1d2b7dee00e90e52700e+431f5505/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5d8642293405ae45c807bfc2078bf22602f8f53a3824b448e321d6b17fa96a5a
3
+ size 198203
neuronxcc-2.19.8089.0+8ab9f450/MODULE_1d2b7dee00e90e52700e+431f5505/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9cc91d3344e6a7acf2ce96d573914b28841011f1bcef7ea879355a6866286c00
3
+ size 3185664
neuronxcc-2.19.8089.0+8ab9f450/MODULE_41d952393bdf990574ea+431f5505/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ "--model-type=transformer -O1 --lnc=1 --internal-hlo2tensorizer-options=--experimental-unsafe-fp8e4m3fn-as-fp8e4m3 --logfile=/tmp/nxd_model/layout_opt/log-neuron-cc.txt"
neuronxcc-2.19.8089.0+8ab9f450/MODULE_41d952393bdf990574ea+431f5505/model.done ADDED
File without changes
neuronxcc-2.19.8089.0+8ab9f450/MODULE_41d952393bdf990574ea+431f5505/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ac477097f210e0dfb28201a42aa4e76415d59b1d5dbb0eb4e4cbafed02228523
3
+ size 154634
neuronxcc-2.19.8089.0+8ab9f450/MODULE_41d952393bdf990574ea+431f5505/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dad7307699d38332a358b4ff0dcc8b21d650f98d20c5bed693d6b8eb29a80916
3
+ size 3144704
neuronxcc-2.19.8089.0+8ab9f450/MODULE_42f2f26b5d67b722528c+431f5505/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ "--model-type=transformer -O1 --lnc=1 --internal-hlo2tensorizer-options=--experimental-unsafe-fp8e4m3fn-as-fp8e4m3 --logfile=/tmp/nxd_model/layout_opt/log-neuron-cc.txt"
neuronxcc-2.19.8089.0+8ab9f450/MODULE_42f2f26b5d67b722528c+431f5505/model.done ADDED
File without changes
neuronxcc-2.19.8089.0+8ab9f450/MODULE_42f2f26b5d67b722528c+431f5505/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2bb5d459df79786becdd9fdd9f89faf2d01dd48e1dc407a6e2bb6561214e499c
3
+ size 154641
neuronxcc-2.19.8089.0+8ab9f450/MODULE_42f2f26b5d67b722528c+431f5505/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ce293a6c1b9514f868bd95699df03c472d7ff9126a2fd602268ea42e6b7f5974
3
+ size 3144704
neuronxcc-2.19.8089.0+8ab9f450/MODULE_5a4f005f17fc9cd3a4ea+a9d440f5/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_model/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
neuronxcc-2.19.8089.0+8ab9f450/MODULE_5a4f005f17fc9cd3a4ea+a9d440f5/model.done ADDED
File without changes
neuronxcc-2.19.8089.0+8ab9f450/MODULE_5a4f005f17fc9cd3a4ea+a9d440f5/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:640be59bb8ab96ddc9429042250579daa625b1975213591f9366e42938417e17
3
+ size 795264
neuronxcc-2.19.8089.0+8ab9f450/MODULE_5a4f005f17fc9cd3a4ea+a9d440f5/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9461b0a955924e73f9d02e72a6b2d8e921a6c8613b5835b277000810e13daf2c
3
+ size 3820544
neuronxcc-2.19.8089.0+8ab9f450/MODULE_5a4f005f17fc9cd3a4ea+a9d440f5/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:46acde2ffd2a6732876b9de805b2590005aeeb01fb31144d006024a8272d5529
3
+ size 3976689
neuronxcc-2.19.8089.0+8ab9f450/MODULE_865d84616d5b312d9a15+ed72d204/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_model/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.19.8089.0+8ab9f450/MODULE_865d84616d5b312d9a15+ed72d204/model.done ADDED
File without changes
neuronxcc-2.19.8089.0+8ab9f450/MODULE_865d84616d5b312d9a15+ed72d204/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a9760d3d54d42bb78c35e72d673363e80136997195c2f33b331cef3264018a84
3
+ size 887761