dacorvo HF Staff commited on
Commit
3f5fb5e
·
verified ·
1 Parent(s): 5ffe9ee

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-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/granite/ibm-granite/granite-3.1-2b-instruct/0f6c4fdc5392f85cc1a9.json +59 -0
  3. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/granite/ibm-granite/granite-3.1-2b-instruct/776cfdcbfedab12abaef.json +59 -0
  4. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/meta-llama/Llama-3.1-8B-Instruct/52a3b7d021f51c90337f.json +63 -0
  5. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/meta-llama/Llama-3.1-8B-Instruct/5fdba651620df09da93d.json +63 -0
  6. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/meta-llama/Llama-3.1-8B-Instruct/daa276345bb9b68e9be5.json +63 -0
  7. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/meta-llama/Llama-3.1-8B-Instruct/e4b573e1a33bbda76243.json +63 -0
  8. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/meta-llama/Llama-3.2-1B-Instruct/ae1848094edb282bfdf1.json +63 -0
  9. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/meta-llama/Meta-Llama-3.1-8B-Instruct/c0cef3ac9ffe05625f39.json +63 -0
  10. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/unsloth/Llama-3.2-1B-Instruct/683e082396a4bd7ab4cb.json +64 -0
  11. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/unsloth/Llama-3.2-1B-Instruct/79f88038b5962d921f3d.json +64 -0
  12. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/unsloth/Llama-3.2-1B-Instruct/8b175859f95b9fa5a1db.json +63 -0
  13. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/unsloth/Llama-3.2-1B-Instruct/93fa413575d5ccc52f58.json +63 -0
  14. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/unsloth/Llama-3.2-1B-Instruct/f453e7e94ffe46947fc0.json +64 -0
  15. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/unsloth/llama-3.2-1B-Instruct/c093ba8faccfd3d9f2c2.json +64 -0
  16. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/phi3/microsoft/Phi-3.5-mini-instruct/86c6ae54af67ddf4b3b8.json +165 -0
  17. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/phi3/microsoft/Phi-3.5-mini-instruct/bb3ccea738a2a8e75fe8.json +165 -0
  18. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/qwen2/Qwen/Qwen2.5-0.5B/53917a90164d81443a08.json +83 -0
  19. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/qwen2/Qwen/Qwen2.5-0.5B/b558da325694139bbaaa.json +83 -0
  20. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/qwen3/Qwen/Qwen3-0.6B/1b21de7b61f3d14bbbff.json +88 -0
  21. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/qwen3/Qwen/Qwen3-0.6B/ad8740a6bddc8f6b87a7.json +88 -0
  22. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/smollm3/HuggingFaceTB/SmolLM3-3B/0165b063d3dfd73ec9af.json +135 -0
  23. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/smollm3/HuggingFaceTB/SmolLM3-3B/71933d2c6e6099113f93.json +135 -0
  24. neuronxcc-2.21.33363.0+82129205/MODULE_110fc80e89006393f738+24129607/compile_flags.json +1 -0
  25. neuronxcc-2.21.33363.0+82129205/MODULE_110fc80e89006393f738+24129607/model.done +0 -0
  26. neuronxcc-2.21.33363.0+82129205/MODULE_110fc80e89006393f738+24129607/model.hlo_module.pb +3 -0
  27. neuronxcc-2.21.33363.0+82129205/MODULE_110fc80e89006393f738+24129607/model.neff +3 -0
  28. neuronxcc-2.21.33363.0+82129205/MODULE_196d1b2148ed8629b154+24129607/compile_flags.json +1 -0
  29. neuronxcc-2.21.33363.0+82129205/MODULE_196d1b2148ed8629b154+24129607/model.done +0 -0
  30. neuronxcc-2.21.33363.0+82129205/MODULE_196d1b2148ed8629b154+24129607/model.hlo_module.pb +3 -0
  31. neuronxcc-2.21.33363.0+82129205/MODULE_196d1b2148ed8629b154+24129607/model.neff +3 -0
  32. neuronxcc-2.21.33363.0+82129205/MODULE_1aa5f4baa9354745d6a6+24129607/compile_flags.json +1 -0
  33. neuronxcc-2.21.33363.0+82129205/MODULE_1aa5f4baa9354745d6a6+24129607/model.done +0 -0
  34. neuronxcc-2.21.33363.0+82129205/MODULE_1aa5f4baa9354745d6a6+24129607/model.hlo_module.pb +3 -0
  35. neuronxcc-2.21.33363.0+82129205/MODULE_1aa5f4baa9354745d6a6+24129607/model.neff +3 -0
  36. neuronxcc-2.21.33363.0+82129205/MODULE_37d321becc90cb687039+24129607/compile_flags.json +1 -0
  37. neuronxcc-2.21.33363.0+82129205/MODULE_37d321becc90cb687039+24129607/model.done +0 -0
  38. neuronxcc-2.21.33363.0+82129205/MODULE_37d321becc90cb687039+24129607/model.hlo_module.pb +3 -0
  39. neuronxcc-2.21.33363.0+82129205/MODULE_37d321becc90cb687039+24129607/model.neff +3 -0
  40. neuronxcc-2.21.33363.0+82129205/MODULE_39061f1efbca2332dc73+24129607/compile_flags.json +1 -0
  41. neuronxcc-2.21.33363.0+82129205/MODULE_39061f1efbca2332dc73+24129607/model.done +0 -0
  42. neuronxcc-2.21.33363.0+82129205/MODULE_39061f1efbca2332dc73+24129607/model.hlo_module.pb +3 -0
  43. neuronxcc-2.21.33363.0+82129205/MODULE_39061f1efbca2332dc73+24129607/model.neff +3 -0
  44. neuronxcc-2.21.33363.0+82129205/MODULE_5cfe268b844f7d2286a5+a02c3a36/compile_flags.json +1 -0
  45. neuronxcc-2.21.33363.0+82129205/MODULE_5cfe268b844f7d2286a5+a02c3a36/model.done +0 -0
  46. neuronxcc-2.21.33363.0+82129205/MODULE_5cfe268b844f7d2286a5+a02c3a36/model.hlo_module.pb +3 -0
  47. neuronxcc-2.21.33363.0+82129205/MODULE_5cfe268b844f7d2286a5+a02c3a36/model.neff +3 -0
  48. neuronxcc-2.21.33363.0+82129205/MODULE_5cfe268b844f7d2286a5+a02c3a36/wrapped_neff.hlo +3 -0
  49. neuronxcc-2.21.33363.0+82129205/MODULE_7c1ab0225123c184780f+24129607/compile_flags.json +1 -0
  50. neuronxcc-2.21.33363.0+82129205/MODULE_7c1ab0225123c184780f+24129607/model.done +0 -0
.gitattributes CHANGED
@@ -15032,3 +15032,19 @@ neuronxcc-2.21.33363.0+82129205/MODULE_d7f7c69ad6f63a27e1a5+a02c3a36/model.neff
15032
  neuronxcc-2.21.33363.0+82129205/MODULE_d7f7c69ad6f63a27e1a5+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
15033
  neuronxcc-2.21.33363.0+82129205/MODULE_09e98045a01eb4a75c24+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
15034
  neuronxcc-2.21.33363.0+82129205/MODULE_09e98045a01eb4a75c24+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
15032
  neuronxcc-2.21.33363.0+82129205/MODULE_d7f7c69ad6f63a27e1a5+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
15033
  neuronxcc-2.21.33363.0+82129205/MODULE_09e98045a01eb4a75c24+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
15034
  neuronxcc-2.21.33363.0+82129205/MODULE_09e98045a01eb4a75c24+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
15035
+ neuronxcc-2.21.33363.0+82129205/MODULE_110fc80e89006393f738+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
15036
+ neuronxcc-2.21.33363.0+82129205/MODULE_196d1b2148ed8629b154+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
15037
+ neuronxcc-2.21.33363.0+82129205/MODULE_1aa5f4baa9354745d6a6+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
15038
+ neuronxcc-2.21.33363.0+82129205/MODULE_37d321becc90cb687039+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
15039
+ neuronxcc-2.21.33363.0+82129205/MODULE_39061f1efbca2332dc73+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
15040
+ neuronxcc-2.21.33363.0+82129205/MODULE_5cfe268b844f7d2286a5+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
15041
+ neuronxcc-2.21.33363.0+82129205/MODULE_5cfe268b844f7d2286a5+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
15042
+ neuronxcc-2.21.33363.0+82129205/MODULE_7c1ab0225123c184780f+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
15043
+ neuronxcc-2.21.33363.0+82129205/MODULE_88399bbf2a34b1e28eee+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
15044
+ neuronxcc-2.21.33363.0+82129205/MODULE_9b8221eb10f6b4eb2f68+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
15045
+ neuronxcc-2.21.33363.0+82129205/MODULE_9b8221eb10f6b4eb2f68+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
15046
+ neuronxcc-2.21.33363.0+82129205/MODULE_a6d912262b31e81edfe6+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
15047
+ neuronxcc-2.21.33363.0+82129205/MODULE_cfd70138eb9722ac2255+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
15048
+ neuronxcc-2.21.33363.0+82129205/MODULE_cfd70138eb9722ac2255+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
15049
+ neuronxcc-2.21.33363.0+82129205/MODULE_dd366124647aeec64074+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
15050
+ neuronxcc-2.21.33363.0+82129205/MODULE_dd366124647aeec64074+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/granite/ibm-granite/granite-3.1-2b-instruct/0f6c4fdc5392f85cc1a9.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "ibm-granite/granite-3.1-2b-instruct",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "GraniteForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.1,
10
+ "attention_multiplier": 0.015625,
11
+ "dtype": "bfloat16",
12
+ "embedding_multiplier": 12.0,
13
+ "hidden_act": "silu",
14
+ "hidden_size": 2048,
15
+ "initializer_range": 0.02,
16
+ "intermediate_size": 8192,
17
+ "logits_scaling": 8.0,
18
+ "max_position_embeddings": 131072,
19
+ "mlp_bias": false,
20
+ "model_type": "granite",
21
+ "neuron": {
22
+ "_serialized_key": "NxDNeuronConfig",
23
+ "batch_size": 1,
24
+ "capacity_factor": null,
25
+ "checkpoint_id": "ibm-granite/granite-3.1-2b-instruct",
26
+ "checkpoint_revision": "bbc2aed595bd38bd770263dc3ab831db9794441d",
27
+ "continuous_batching": false,
28
+ "ep_degree": 1,
29
+ "fused_qkv": true,
30
+ "glu_mlp": true,
31
+ "local_ranks_size": 2,
32
+ "max_batch_size": 1,
33
+ "max_context_length": 8192,
34
+ "max_topk": 256,
35
+ "n_active_tokens": 8192,
36
+ "neuronxcc_version": "2.21.33363.0+82129205",
37
+ "on_device_sampling": true,
38
+ "optimum_neuron_version": "0.4.4.dev2",
39
+ "output_logits": false,
40
+ "pp_degree": 1,
41
+ "sequence_length": 8192,
42
+ "sequence_parallel_enabled": true,
43
+ "speculation_length": 0,
44
+ "start_rank_id": 0,
45
+ "target": "trn1",
46
+ "torch_dtype": "bfloat16",
47
+ "tp_degree": 2
48
+ },
49
+ "num_attention_heads": 32,
50
+ "num_hidden_layers": 40,
51
+ "num_key_value_heads": 8,
52
+ "residual_multiplier": 0.22,
53
+ "rms_norm_eps": 1e-05,
54
+ "rope_scaling": null,
55
+ "rope_theta": 5000000.0,
56
+ "tie_word_embeddings": true,
57
+ "use_cache": true,
58
+ "vocab_size": 49155
59
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/granite/ibm-granite/granite-3.1-2b-instruct/776cfdcbfedab12abaef.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "ibm-granite/granite-3.1-2b-instruct",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "GraniteForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.1,
10
+ "attention_multiplier": 0.015625,
11
+ "dtype": "bfloat16",
12
+ "embedding_multiplier": 12.0,
13
+ "hidden_act": "silu",
14
+ "hidden_size": 2048,
15
+ "initializer_range": 0.02,
16
+ "intermediate_size": 8192,
17
+ "logits_scaling": 8.0,
18
+ "max_position_embeddings": 131072,
19
+ "mlp_bias": false,
20
+ "model_type": "granite",
21
+ "neuron": {
22
+ "_serialized_key": "NxDNeuronConfig",
23
+ "batch_size": 4,
24
+ "capacity_factor": null,
25
+ "checkpoint_id": "ibm-granite/granite-3.1-2b-instruct",
26
+ "checkpoint_revision": "bbc2aed595bd38bd770263dc3ab831db9794441d",
27
+ "continuous_batching": true,
28
+ "ep_degree": 1,
29
+ "fused_qkv": true,
30
+ "glu_mlp": true,
31
+ "local_ranks_size": 2,
32
+ "max_batch_size": 4,
33
+ "max_context_length": 4096,
34
+ "max_topk": 256,
35
+ "n_active_tokens": 4096,
36
+ "neuronxcc_version": "2.21.33363.0+82129205",
37
+ "on_device_sampling": true,
38
+ "optimum_neuron_version": "0.4.4.dev2",
39
+ "output_logits": false,
40
+ "pp_degree": 1,
41
+ "sequence_length": 4096,
42
+ "sequence_parallel_enabled": false,
43
+ "speculation_length": 0,
44
+ "start_rank_id": 0,
45
+ "target": "trn1",
46
+ "torch_dtype": "bfloat16",
47
+ "tp_degree": 2
48
+ },
49
+ "num_attention_heads": 32,
50
+ "num_hidden_layers": 40,
51
+ "num_key_value_heads": 8,
52
+ "residual_multiplier": 0.22,
53
+ "rms_norm_eps": 1e-05,
54
+ "rope_scaling": null,
55
+ "rope_theta": 5000000.0,
56
+ "tie_word_embeddings": true,
57
+ "use_cache": true,
58
+ "vocab_size": 49155
59
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/meta-llama/Llama-3.1-8B-Instruct/52a3b7d021f51c90337f.json ADDED
@@ -0,0 +1,63 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "meta-llama/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": 32,
22
+ "capacity_factor": null,
23
+ "checkpoint_id": "meta-llama/Llama-3.1-8B-Instruct",
24
+ "checkpoint_revision": "0e9e39f249a16976918f6564b8830bc894c89659",
25
+ "continuous_batching": true,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 2,
30
+ "max_batch_size": 32,
31
+ "max_context_length": 4096,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 4096,
34
+ "neuronxcc_version": "2.21.33363.0+82129205",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.4.dev2",
37
+ "output_logits": false,
38
+ "pp_degree": 1,
39
+ "sequence_length": 4096,
40
+ "sequence_parallel_enabled": false,
41
+ "speculation_length": 0,
42
+ "start_rank_id": 0,
43
+ "target": "trn1",
44
+ "torch_dtype": "bfloat16",
45
+ "tp_degree": 2
46
+ },
47
+ "num_attention_heads": 32,
48
+ "num_hidden_layers": 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
+ "use_cache": true,
62
+ "vocab_size": 128256
63
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/meta-llama/Llama-3.1-8B-Instruct/5fdba651620df09da93d.json ADDED
@@ -0,0 +1,63 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "meta-llama/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": 32,
22
+ "capacity_factor": null,
23
+ "checkpoint_id": "meta-llama/Llama-3.1-8B-Instruct",
24
+ "checkpoint_revision": "0e9e39f249a16976918f6564b8830bc894c89659",
25
+ "continuous_batching": true,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 8,
30
+ "max_batch_size": 32,
31
+ "max_context_length": 4096,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 4096,
34
+ "neuronxcc_version": "2.21.33363.0+82129205",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.4.dev2",
37
+ "output_logits": false,
38
+ "pp_degree": 1,
39
+ "sequence_length": 4096,
40
+ "sequence_parallel_enabled": false,
41
+ "speculation_length": 0,
42
+ "start_rank_id": 0,
43
+ "target": "trn1",
44
+ "torch_dtype": "bfloat16",
45
+ "tp_degree": 8
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
+ "use_cache": true,
62
+ "vocab_size": 128256
63
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/meta-llama/Llama-3.1-8B-Instruct/daa276345bb9b68e9be5.json ADDED
@@ -0,0 +1,63 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "meta-llama/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": 32,
22
+ "capacity_factor": null,
23
+ "checkpoint_id": "meta-llama/Llama-3.1-8B-Instruct",
24
+ "checkpoint_revision": "0e9e39f249a16976918f6564b8830bc894c89659",
25
+ "continuous_batching": true,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 8,
30
+ "max_batch_size": 32,
31
+ "max_context_length": 4096,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 4096,
34
+ "neuronxcc_version": "2.21.33363.0+82129205",
35
+ "on_device_sampling": false,
36
+ "optimum_neuron_version": "0.4.4.dev2",
37
+ "output_logits": false,
38
+ "pp_degree": 1,
39
+ "sequence_length": 4096,
40
+ "sequence_parallel_enabled": false,
41
+ "speculation_length": 0,
42
+ "start_rank_id": 0,
43
+ "target": "trn1",
44
+ "torch_dtype": "bfloat16",
45
+ "tp_degree": 8
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
+ "use_cache": true,
62
+ "vocab_size": 128256
63
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/meta-llama/Llama-3.1-8B-Instruct/e4b573e1a33bbda76243.json ADDED
@@ -0,0 +1,63 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "meta-llama/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": 32,
22
+ "capacity_factor": null,
23
+ "checkpoint_id": "meta-llama/Llama-3.1-8B-Instruct",
24
+ "checkpoint_revision": "0e9e39f249a16976918f6564b8830bc894c89659",
25
+ "continuous_batching": true,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 2,
30
+ "max_batch_size": 32,
31
+ "max_context_length": 4096,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 4096,
34
+ "neuronxcc_version": "2.21.33363.0+82129205",
35
+ "on_device_sampling": false,
36
+ "optimum_neuron_version": "0.4.4.dev2",
37
+ "output_logits": false,
38
+ "pp_degree": 1,
39
+ "sequence_length": 4096,
40
+ "sequence_parallel_enabled": false,
41
+ "speculation_length": 0,
42
+ "start_rank_id": 0,
43
+ "target": "trn1",
44
+ "torch_dtype": "bfloat16",
45
+ "tp_degree": 2
46
+ },
47
+ "num_attention_heads": 32,
48
+ "num_hidden_layers": 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
+ "use_cache": true,
62
+ "vocab_size": 128256
63
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/meta-llama/Llama-3.2-1B-Instruct/ae1848094edb282bfdf1.json ADDED
@@ -0,0 +1,63 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "meta-llama/Llama-3.2-1B-Instruct",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "dtype": "bfloat16",
11
+ "head_dim": 64,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 2048,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 8192,
16
+ "max_position_embeddings": 131072,
17
+ "mlp_bias": false,
18
+ "model_type": "llama",
19
+ "neuron": {
20
+ "_serialized_key": "NxDNeuronConfig",
21
+ "batch_size": 1,
22
+ "capacity_factor": null,
23
+ "checkpoint_id": "meta-llama/Llama-3.2-1B-Instruct",
24
+ "checkpoint_revision": "9213176726f574b556790deb65791e0c5aa438b6",
25
+ "continuous_batching": false,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 2,
30
+ "max_batch_size": 1,
31
+ "max_context_length": 8192,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 8192,
34
+ "neuronxcc_version": "2.21.33363.0+82129205",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.4.dev2",
37
+ "output_logits": false,
38
+ "pp_degree": 1,
39
+ "sequence_length": 8192,
40
+ "sequence_parallel_enabled": true,
41
+ "speculation_length": 0,
42
+ "start_rank_id": 0,
43
+ "target": "trn1",
44
+ "torch_dtype": "bfloat16",
45
+ "tp_degree": 2
46
+ },
47
+ "num_attention_heads": 32,
48
+ "num_hidden_layers": 16,
49
+ "num_key_value_heads": 8,
50
+ "pretraining_tp": 1,
51
+ "rms_norm_eps": 1e-05,
52
+ "rope_scaling": {
53
+ "factor": 32.0,
54
+ "high_freq_factor": 4.0,
55
+ "low_freq_factor": 1.0,
56
+ "original_max_position_embeddings": 8192,
57
+ "rope_type": "llama3"
58
+ },
59
+ "rope_theta": 500000.0,
60
+ "tie_word_embeddings": true,
61
+ "use_cache": true,
62
+ "vocab_size": 128256
63
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/meta-llama/Meta-Llama-3.1-8B-Instruct/c0cef3ac9ffe05625f39.json ADDED
@@ -0,0 +1,63 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "meta-llama/Meta-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": 32,
22
+ "capacity_factor": null,
23
+ "checkpoint_id": "meta-llama/Meta-Llama-3.1-8B-Instruct",
24
+ "checkpoint_revision": "0e9e39f249a16976918f6564b8830bc894c89659",
25
+ "continuous_batching": true,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 8,
30
+ "max_batch_size": 32,
31
+ "max_context_length": 4096,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 4096,
34
+ "neuronxcc_version": "2.21.33363.0+82129205",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.4.dev2",
37
+ "output_logits": false,
38
+ "pp_degree": 1,
39
+ "sequence_length": 4096,
40
+ "sequence_parallel_enabled": false,
41
+ "speculation_length": 0,
42
+ "start_rank_id": 0,
43
+ "target": "trn1",
44
+ "torch_dtype": "bfloat16",
45
+ "tp_degree": 8
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
+ "use_cache": true,
62
+ "vocab_size": 128256
63
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/unsloth/Llama-3.2-1B-Instruct/683e082396a4bd7ab4cb.json ADDED
@@ -0,0 +1,64 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "unsloth/Llama-3.2-1B-Instruct",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "dtype": "bfloat16",
11
+ "head_dim": 64,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 2048,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 8192,
16
+ "max_position_embeddings": 131072,
17
+ "mlp_bias": false,
18
+ "model_type": "llama",
19
+ "neuron": {
20
+ "_serialized_key": "NxDNeuronConfig",
21
+ "batch_size": 4,
22
+ "capacity_factor": null,
23
+ "checkpoint_id": "unsloth/Llama-3.2-1B-Instruct",
24
+ "checkpoint_revision": "5a8abab4a5d6f164389b1079fb721cfab8d7126c",
25
+ "continuous_batching": true,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 2,
30
+ "max_batch_size": 4,
31
+ "max_context_length": 4096,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 4096,
34
+ "neuronxcc_version": "2.21.33363.0+82129205",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.4.dev2",
37
+ "output_logits": false,
38
+ "pp_degree": 1,
39
+ "sequence_length": 4096,
40
+ "sequence_parallel_enabled": false,
41
+ "speculation_length": 0,
42
+ "start_rank_id": 0,
43
+ "target": "trn1",
44
+ "torch_dtype": "bfloat16",
45
+ "tp_degree": 2
46
+ },
47
+ "num_attention_heads": 32,
48
+ "num_hidden_layers": 16,
49
+ "num_key_value_heads": 8,
50
+ "pretraining_tp": 1,
51
+ "rms_norm_eps": 1e-05,
52
+ "rope_scaling": {
53
+ "factor": 32.0,
54
+ "high_freq_factor": 4.0,
55
+ "low_freq_factor": 1.0,
56
+ "original_max_position_embeddings": 8192,
57
+ "rope_type": "llama3"
58
+ },
59
+ "rope_theta": 500000.0,
60
+ "tie_word_embeddings": true,
61
+ "unsloth_fixed": true,
62
+ "use_cache": true,
63
+ "vocab_size": 128256
64
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/unsloth/Llama-3.2-1B-Instruct/79f88038b5962d921f3d.json ADDED
@@ -0,0 +1,64 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "unsloth/Llama-3.2-1B-Instruct",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "dtype": "bfloat16",
11
+ "head_dim": 64,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 2048,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 8192,
16
+ "max_position_embeddings": 131072,
17
+ "mlp_bias": false,
18
+ "model_type": "llama",
19
+ "neuron": {
20
+ "_serialized_key": "NxDNeuronConfig",
21
+ "batch_size": 1,
22
+ "capacity_factor": null,
23
+ "checkpoint_id": "unsloth/Llama-3.2-1B-Instruct",
24
+ "checkpoint_revision": "5a8abab4a5d6f164389b1079fb721cfab8d7126c",
25
+ "continuous_batching": false,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 2,
30
+ "max_batch_size": 1,
31
+ "max_context_length": 8192,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 8192,
34
+ "neuronxcc_version": "2.21.33363.0+82129205",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.4.dev2",
37
+ "output_logits": false,
38
+ "pp_degree": 1,
39
+ "sequence_length": 8192,
40
+ "sequence_parallel_enabled": true,
41
+ "speculation_length": 0,
42
+ "start_rank_id": 0,
43
+ "target": "trn1",
44
+ "torch_dtype": "bfloat16",
45
+ "tp_degree": 2
46
+ },
47
+ "num_attention_heads": 32,
48
+ "num_hidden_layers": 16,
49
+ "num_key_value_heads": 8,
50
+ "pretraining_tp": 1,
51
+ "rms_norm_eps": 1e-05,
52
+ "rope_scaling": {
53
+ "factor": 32.0,
54
+ "high_freq_factor": 4.0,
55
+ "low_freq_factor": 1.0,
56
+ "original_max_position_embeddings": 8192,
57
+ "rope_type": "llama3"
58
+ },
59
+ "rope_theta": 500000.0,
60
+ "tie_word_embeddings": true,
61
+ "unsloth_fixed": true,
62
+ "use_cache": true,
63
+ "vocab_size": 128256
64
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/unsloth/Llama-3.2-1B-Instruct/8b175859f95b9fa5a1db.json ADDED
@@ -0,0 +1,63 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "unsloth/Llama-3.2-1B-Instruct",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "dtype": "bfloat16",
11
+ "head_dim": 64,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 2048,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 8192,
16
+ "max_position_embeddings": 131072,
17
+ "mlp_bias": false,
18
+ "model_type": "llama",
19
+ "neuron": {
20
+ "_serialized_key": "NxDNeuronConfig",
21
+ "batch_size": 4,
22
+ "capacity_factor": null,
23
+ "checkpoint_id": "unsloth/Llama-3.2-1B-Instruct",
24
+ "checkpoint_revision": "5a8abab4a5d6f164389b1079fb721cfab8d7126c",
25
+ "continuous_batching": true,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 2,
30
+ "max_batch_size": 4,
31
+ "max_context_length": 4096,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 4096,
34
+ "neuronxcc_version": "2.21.33363.0+82129205",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.4.dev2",
37
+ "output_logits": false,
38
+ "pp_degree": 1,
39
+ "sequence_length": 4096,
40
+ "speculation_length": 0,
41
+ "start_rank_id": 0,
42
+ "target": "trn1",
43
+ "torch_dtype": "bfloat16",
44
+ "tp_degree": 2
45
+ },
46
+ "num_attention_heads": 32,
47
+ "num_hidden_layers": 16,
48
+ "num_key_value_heads": 8,
49
+ "pretraining_tp": 1,
50
+ "rms_norm_eps": 1e-05,
51
+ "rope_scaling": {
52
+ "factor": 32.0,
53
+ "high_freq_factor": 4.0,
54
+ "low_freq_factor": 1.0,
55
+ "original_max_position_embeddings": 8192,
56
+ "rope_type": "llama3"
57
+ },
58
+ "rope_theta": 500000.0,
59
+ "tie_word_embeddings": true,
60
+ "unsloth_fixed": true,
61
+ "use_cache": true,
62
+ "vocab_size": 128256
63
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/unsloth/Llama-3.2-1B-Instruct/93fa413575d5ccc52f58.json ADDED
@@ -0,0 +1,63 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "unsloth/Llama-3.2-1B-Instruct",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "dtype": "bfloat16",
11
+ "head_dim": 64,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 2048,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 8192,
16
+ "max_position_embeddings": 131072,
17
+ "mlp_bias": false,
18
+ "model_type": "llama",
19
+ "neuron": {
20
+ "_serialized_key": "NxDNeuronConfig",
21
+ "batch_size": 1,
22
+ "capacity_factor": null,
23
+ "checkpoint_id": "unsloth/Llama-3.2-1B-Instruct",
24
+ "checkpoint_revision": "5a8abab4a5d6f164389b1079fb721cfab8d7126c",
25
+ "continuous_batching": false,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 2,
30
+ "max_batch_size": 1,
31
+ "max_context_length": 8192,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 8192,
34
+ "neuronxcc_version": "2.21.33363.0+82129205",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.4.dev2",
37
+ "output_logits": false,
38
+ "pp_degree": 1,
39
+ "sequence_length": 8192,
40
+ "speculation_length": 0,
41
+ "start_rank_id": 0,
42
+ "target": "trn1",
43
+ "torch_dtype": "bfloat16",
44
+ "tp_degree": 2
45
+ },
46
+ "num_attention_heads": 32,
47
+ "num_hidden_layers": 16,
48
+ "num_key_value_heads": 8,
49
+ "pretraining_tp": 1,
50
+ "rms_norm_eps": 1e-05,
51
+ "rope_scaling": {
52
+ "factor": 32.0,
53
+ "high_freq_factor": 4.0,
54
+ "low_freq_factor": 1.0,
55
+ "original_max_position_embeddings": 8192,
56
+ "rope_type": "llama3"
57
+ },
58
+ "rope_theta": 500000.0,
59
+ "tie_word_embeddings": true,
60
+ "unsloth_fixed": true,
61
+ "use_cache": true,
62
+ "vocab_size": 128256
63
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/unsloth/Llama-3.2-1B-Instruct/f453e7e94ffe46947fc0.json ADDED
@@ -0,0 +1,64 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "unsloth/Llama-3.2-1B-Instruct",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "dtype": "bfloat16",
11
+ "head_dim": 64,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 2048,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 8192,
16
+ "max_position_embeddings": 131072,
17
+ "mlp_bias": false,
18
+ "model_type": "llama",
19
+ "neuron": {
20
+ "_serialized_key": "NxDNeuronConfig",
21
+ "batch_size": 1,
22
+ "capacity_factor": null,
23
+ "checkpoint_id": "unsloth/Llama-3.2-1B-Instruct",
24
+ "checkpoint_revision": "5a8abab4a5d6f164389b1079fb721cfab8d7126c",
25
+ "continuous_batching": false,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 2,
30
+ "max_batch_size": 1,
31
+ "max_context_length": 8192,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 8192,
34
+ "neuronxcc_version": "2.21.33363.0+82129205",
35
+ "on_device_sampling": false,
36
+ "optimum_neuron_version": "0.4.4.dev2",
37
+ "output_logits": false,
38
+ "pp_degree": 1,
39
+ "sequence_length": 8192,
40
+ "sequence_parallel_enabled": true,
41
+ "speculation_length": 0,
42
+ "start_rank_id": 0,
43
+ "target": "trn1",
44
+ "torch_dtype": "bfloat16",
45
+ "tp_degree": 2
46
+ },
47
+ "num_attention_heads": 32,
48
+ "num_hidden_layers": 16,
49
+ "num_key_value_heads": 8,
50
+ "pretraining_tp": 1,
51
+ "rms_norm_eps": 1e-05,
52
+ "rope_scaling": {
53
+ "factor": 32.0,
54
+ "high_freq_factor": 4.0,
55
+ "low_freq_factor": 1.0,
56
+ "original_max_position_embeddings": 8192,
57
+ "rope_type": "llama3"
58
+ },
59
+ "rope_theta": 500000.0,
60
+ "tie_word_embeddings": true,
61
+ "unsloth_fixed": true,
62
+ "use_cache": true,
63
+ "vocab_size": 128256
64
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/unsloth/llama-3.2-1B-Instruct/c093ba8faccfd3d9f2c2.json ADDED
@@ -0,0 +1,64 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "unsloth/llama-3.2-1B-Instruct",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "dtype": "bfloat16",
11
+ "head_dim": 64,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 2048,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 8192,
16
+ "max_position_embeddings": 131072,
17
+ "mlp_bias": false,
18
+ "model_type": "llama",
19
+ "neuron": {
20
+ "_serialized_key": "NxDNeuronConfig",
21
+ "batch_size": 4,
22
+ "capacity_factor": null,
23
+ "checkpoint_id": "unsloth/llama-3.2-1B-Instruct",
24
+ "checkpoint_revision": "5a8abab4a5d6f164389b1079fb721cfab8d7126c",
25
+ "continuous_batching": true,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 2,
30
+ "max_batch_size": 4,
31
+ "max_context_length": 4096,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 4096,
34
+ "neuronxcc_version": "2.21.33363.0+82129205",
35
+ "on_device_sampling": false,
36
+ "optimum_neuron_version": "0.4.4.dev2",
37
+ "output_logits": false,
38
+ "pp_degree": 1,
39
+ "sequence_length": 4096,
40
+ "sequence_parallel_enabled": false,
41
+ "speculation_length": 0,
42
+ "start_rank_id": 0,
43
+ "target": "trn1",
44
+ "torch_dtype": "bfloat16",
45
+ "tp_degree": 2
46
+ },
47
+ "num_attention_heads": 32,
48
+ "num_hidden_layers": 16,
49
+ "num_key_value_heads": 8,
50
+ "pretraining_tp": 1,
51
+ "rms_norm_eps": 1e-05,
52
+ "rope_scaling": {
53
+ "factor": 32.0,
54
+ "high_freq_factor": 4.0,
55
+ "low_freq_factor": 1.0,
56
+ "original_max_position_embeddings": 8192,
57
+ "rope_type": "llama3"
58
+ },
59
+ "rope_theta": 500000.0,
60
+ "tie_word_embeddings": true,
61
+ "unsloth_fixed": true,
62
+ "use_cache": true,
63
+ "vocab_size": 128256
64
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/phi3/microsoft/Phi-3.5-mini-instruct/86c6ae54af67ddf4b3b8.json ADDED
@@ -0,0 +1,165 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "microsoft/Phi-3.5-mini-instruct",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Phi3ForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "auto_map": {
11
+ "AutoConfig": "configuration_phi3.Phi3Config",
12
+ "AutoModelForCausalLM": "modeling_phi3.Phi3ForCausalLM"
13
+ },
14
+ "dtype": "bfloat16",
15
+ "embd_pdrop": 0.0,
16
+ "hidden_act": "silu",
17
+ "hidden_size": 3072,
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 8192,
20
+ "max_position_embeddings": 131072,
21
+ "model_type": "phi3",
22
+ "neuron": {
23
+ "_serialized_key": "NxDNeuronConfig",
24
+ "batch_size": 1,
25
+ "capacity_factor": null,
26
+ "checkpoint_id": "microsoft/Phi-3.5-mini-instruct",
27
+ "checkpoint_revision": "2fe192450127e6a83f7441aef6e3ca586c338b77",
28
+ "continuous_batching": false,
29
+ "ep_degree": 1,
30
+ "fused_qkv": true,
31
+ "glu_mlp": true,
32
+ "local_ranks_size": 2,
33
+ "max_batch_size": 1,
34
+ "max_context_length": 8192,
35
+ "max_topk": 256,
36
+ "n_active_tokens": 8192,
37
+ "neuronxcc_version": "2.21.33363.0+82129205",
38
+ "on_device_sampling": true,
39
+ "optimum_neuron_version": "0.4.4.dev2",
40
+ "output_logits": false,
41
+ "pp_degree": 1,
42
+ "sequence_length": 8192,
43
+ "sequence_parallel_enabled": true,
44
+ "speculation_length": 0,
45
+ "start_rank_id": 0,
46
+ "target": "trn1",
47
+ "torch_dtype": "bfloat16",
48
+ "tp_degree": 2
49
+ },
50
+ "num_attention_heads": 32,
51
+ "num_hidden_layers": 32,
52
+ "num_key_value_heads": 32,
53
+ "original_max_position_embeddings": 4096,
54
+ "partial_rotary_factor": 1.0,
55
+ "resid_pdrop": 0.0,
56
+ "rms_norm_eps": 1e-05,
57
+ "rope_scaling": {
58
+ "long_factor": [
59
+ 1.0800000429153442,
60
+ 1.1100000143051147,
61
+ 1.1399999856948853,
62
+ 1.340000033378601,
63
+ 1.5899999141693115,
64
+ 1.600000023841858,
65
+ 1.6200000047683716,
66
+ 2.620000123977661,
67
+ 3.2300000190734863,
68
+ 3.2300000190734863,
69
+ 4.789999961853027,
70
+ 7.400000095367432,
71
+ 7.700000286102295,
72
+ 9.09000015258789,
73
+ 12.199999809265137,
74
+ 17.670000076293945,
75
+ 24.46000099182129,
76
+ 28.57000160217285,
77
+ 30.420001983642578,
78
+ 30.840002059936523,
79
+ 32.590003967285156,
80
+ 32.93000411987305,
81
+ 42.320003509521484,
82
+ 44.96000289916992,
83
+ 50.340003967285156,
84
+ 50.45000457763672,
85
+ 57.55000305175781,
86
+ 57.93000411987305,
87
+ 58.21000289916992,
88
+ 60.1400032043457,
89
+ 62.61000442504883,
90
+ 62.62000274658203,
91
+ 62.71000289916992,
92
+ 63.1400032043457,
93
+ 63.1400032043457,
94
+ 63.77000427246094,
95
+ 63.93000411987305,
96
+ 63.96000289916992,
97
+ 63.970001220703125,
98
+ 64.02999877929688,
99
+ 64.06999969482422,
100
+ 64.08000183105469,
101
+ 64.12000274658203,
102
+ 64.41000366210938,
103
+ 64.4800033569336,
104
+ 64.51000213623047,
105
+ 64.52999877929688,
106
+ 64.83999633789062
107
+ ],
108
+ "short_factor": [
109
+ 1.0,
110
+ 1.0199999809265137,
111
+ 1.0299999713897705,
112
+ 1.0299999713897705,
113
+ 1.0499999523162842,
114
+ 1.0499999523162842,
115
+ 1.0499999523162842,
116
+ 1.0499999523162842,
117
+ 1.0499999523162842,
118
+ 1.0699999332427979,
119
+ 1.0999999046325684,
120
+ 1.1099998950958252,
121
+ 1.1599998474121094,
122
+ 1.1599998474121094,
123
+ 1.1699998378753662,
124
+ 1.2899998426437378,
125
+ 1.339999794960022,
126
+ 1.679999828338623,
127
+ 1.7899998426437378,
128
+ 1.8199998140335083,
129
+ 1.8499997854232788,
130
+ 1.8799997568130493,
131
+ 1.9099997282028198,
132
+ 1.9399996995925903,
133
+ 1.9899996519088745,
134
+ 2.0199997425079346,
135
+ 2.0199997425079346,
136
+ 2.0199997425079346,
137
+ 2.0199997425079346,
138
+ 2.0199997425079346,
139
+ 2.0199997425079346,
140
+ 2.0299997329711914,
141
+ 2.0299997329711914,
142
+ 2.0299997329711914,
143
+ 2.0299997329711914,
144
+ 2.0299997329711914,
145
+ 2.0299997329711914,
146
+ 2.0299997329711914,
147
+ 2.0299997329711914,
148
+ 2.0299997329711914,
149
+ 2.0799996852874756,
150
+ 2.0899996757507324,
151
+ 2.189999580383301,
152
+ 2.2199995517730713,
153
+ 2.5899994373321533,
154
+ 2.729999542236328,
155
+ 2.749999523162842,
156
+ 2.8399994373321533
157
+ ],
158
+ "type": "longrope"
159
+ },
160
+ "rope_theta": 10000.0,
161
+ "sliding_window": 262144,
162
+ "tie_word_embeddings": false,
163
+ "use_cache": true,
164
+ "vocab_size": 32064
165
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/phi3/microsoft/Phi-3.5-mini-instruct/bb3ccea738a2a8e75fe8.json ADDED
@@ -0,0 +1,165 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "microsoft/Phi-3.5-mini-instruct",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Phi3ForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "auto_map": {
11
+ "AutoConfig": "configuration_phi3.Phi3Config",
12
+ "AutoModelForCausalLM": "modeling_phi3.Phi3ForCausalLM"
13
+ },
14
+ "dtype": "bfloat16",
15
+ "embd_pdrop": 0.0,
16
+ "hidden_act": "silu",
17
+ "hidden_size": 3072,
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 8192,
20
+ "max_position_embeddings": 131072,
21
+ "model_type": "phi3",
22
+ "neuron": {
23
+ "_serialized_key": "NxDNeuronConfig",
24
+ "batch_size": 4,
25
+ "capacity_factor": null,
26
+ "checkpoint_id": "microsoft/Phi-3.5-mini-instruct",
27
+ "checkpoint_revision": "2fe192450127e6a83f7441aef6e3ca586c338b77",
28
+ "continuous_batching": true,
29
+ "ep_degree": 1,
30
+ "fused_qkv": true,
31
+ "glu_mlp": true,
32
+ "local_ranks_size": 2,
33
+ "max_batch_size": 4,
34
+ "max_context_length": 4096,
35
+ "max_topk": 256,
36
+ "n_active_tokens": 4096,
37
+ "neuronxcc_version": "2.21.33363.0+82129205",
38
+ "on_device_sampling": true,
39
+ "optimum_neuron_version": "0.4.4.dev2",
40
+ "output_logits": false,
41
+ "pp_degree": 1,
42
+ "sequence_length": 4096,
43
+ "sequence_parallel_enabled": false,
44
+ "speculation_length": 0,
45
+ "start_rank_id": 0,
46
+ "target": "trn1",
47
+ "torch_dtype": "bfloat16",
48
+ "tp_degree": 2
49
+ },
50
+ "num_attention_heads": 32,
51
+ "num_hidden_layers": 32,
52
+ "num_key_value_heads": 32,
53
+ "original_max_position_embeddings": 4096,
54
+ "partial_rotary_factor": 1.0,
55
+ "resid_pdrop": 0.0,
56
+ "rms_norm_eps": 1e-05,
57
+ "rope_scaling": {
58
+ "long_factor": [
59
+ 1.0800000429153442,
60
+ 1.1100000143051147,
61
+ 1.1399999856948853,
62
+ 1.340000033378601,
63
+ 1.5899999141693115,
64
+ 1.600000023841858,
65
+ 1.6200000047683716,
66
+ 2.620000123977661,
67
+ 3.2300000190734863,
68
+ 3.2300000190734863,
69
+ 4.789999961853027,
70
+ 7.400000095367432,
71
+ 7.700000286102295,
72
+ 9.09000015258789,
73
+ 12.199999809265137,
74
+ 17.670000076293945,
75
+ 24.46000099182129,
76
+ 28.57000160217285,
77
+ 30.420001983642578,
78
+ 30.840002059936523,
79
+ 32.590003967285156,
80
+ 32.93000411987305,
81
+ 42.320003509521484,
82
+ 44.96000289916992,
83
+ 50.340003967285156,
84
+ 50.45000457763672,
85
+ 57.55000305175781,
86
+ 57.93000411987305,
87
+ 58.21000289916992,
88
+ 60.1400032043457,
89
+ 62.61000442504883,
90
+ 62.62000274658203,
91
+ 62.71000289916992,
92
+ 63.1400032043457,
93
+ 63.1400032043457,
94
+ 63.77000427246094,
95
+ 63.93000411987305,
96
+ 63.96000289916992,
97
+ 63.970001220703125,
98
+ 64.02999877929688,
99
+ 64.06999969482422,
100
+ 64.08000183105469,
101
+ 64.12000274658203,
102
+ 64.41000366210938,
103
+ 64.4800033569336,
104
+ 64.51000213623047,
105
+ 64.52999877929688,
106
+ 64.83999633789062
107
+ ],
108
+ "short_factor": [
109
+ 1.0,
110
+ 1.0199999809265137,
111
+ 1.0299999713897705,
112
+ 1.0299999713897705,
113
+ 1.0499999523162842,
114
+ 1.0499999523162842,
115
+ 1.0499999523162842,
116
+ 1.0499999523162842,
117
+ 1.0499999523162842,
118
+ 1.0699999332427979,
119
+ 1.0999999046325684,
120
+ 1.1099998950958252,
121
+ 1.1599998474121094,
122
+ 1.1599998474121094,
123
+ 1.1699998378753662,
124
+ 1.2899998426437378,
125
+ 1.339999794960022,
126
+ 1.679999828338623,
127
+ 1.7899998426437378,
128
+ 1.8199998140335083,
129
+ 1.8499997854232788,
130
+ 1.8799997568130493,
131
+ 1.9099997282028198,
132
+ 1.9399996995925903,
133
+ 1.9899996519088745,
134
+ 2.0199997425079346,
135
+ 2.0199997425079346,
136
+ 2.0199997425079346,
137
+ 2.0199997425079346,
138
+ 2.0199997425079346,
139
+ 2.0199997425079346,
140
+ 2.0299997329711914,
141
+ 2.0299997329711914,
142
+ 2.0299997329711914,
143
+ 2.0299997329711914,
144
+ 2.0299997329711914,
145
+ 2.0299997329711914,
146
+ 2.0299997329711914,
147
+ 2.0299997329711914,
148
+ 2.0299997329711914,
149
+ 2.0799996852874756,
150
+ 2.0899996757507324,
151
+ 2.189999580383301,
152
+ 2.2199995517730713,
153
+ 2.5899994373321533,
154
+ 2.729999542236328,
155
+ 2.749999523162842,
156
+ 2.8399994373321533
157
+ ],
158
+ "type": "longrope"
159
+ },
160
+ "rope_theta": 10000.0,
161
+ "sliding_window": 262144,
162
+ "tie_word_embeddings": false,
163
+ "use_cache": true,
164
+ "vocab_size": 32064
165
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/qwen2/Qwen/Qwen2.5-0.5B/53917a90164d81443a08.json ADDED
@@ -0,0 +1,83 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "Qwen/Qwen2.5-0.5B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Qwen2ForCausalLM"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "dtype": "bfloat16",
10
+ "hidden_act": "silu",
11
+ "hidden_size": 896,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 4864,
14
+ "layer_types": [
15
+ "full_attention",
16
+ "full_attention",
17
+ "full_attention",
18
+ "full_attention",
19
+ "full_attention",
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention"
39
+ ],
40
+ "max_position_embeddings": 32768,
41
+ "max_window_layers": 24,
42
+ "model_type": "qwen2",
43
+ "neuron": {
44
+ "_serialized_key": "NxDNeuronConfig",
45
+ "batch_size": 4,
46
+ "capacity_factor": null,
47
+ "checkpoint_id": "Qwen/Qwen2.5-0.5B",
48
+ "checkpoint_revision": "060db6499f32faf8b98477b0a26969ef7d8b9987",
49
+ "continuous_batching": true,
50
+ "ep_degree": 1,
51
+ "fused_qkv": false,
52
+ "glu_mlp": true,
53
+ "local_ranks_size": 2,
54
+ "max_batch_size": 4,
55
+ "max_context_length": 4096,
56
+ "max_topk": 256,
57
+ "n_active_tokens": 4096,
58
+ "neuronxcc_version": "2.21.33363.0+82129205",
59
+ "on_device_sampling": false,
60
+ "optimum_neuron_version": "0.4.4.dev2",
61
+ "output_logits": false,
62
+ "pp_degree": 1,
63
+ "sequence_length": 4096,
64
+ "sequence_parallel_enabled": false,
65
+ "speculation_length": 0,
66
+ "start_rank_id": 0,
67
+ "target": "trn1",
68
+ "torch_dtype": "bfloat16",
69
+ "tp_degree": 2
70
+ },
71
+ "num_attention_heads": 14,
72
+ "num_hidden_layers": 24,
73
+ "num_key_value_heads": 2,
74
+ "rms_norm_eps": 1e-06,
75
+ "rope_scaling": null,
76
+ "rope_theta": 1000000.0,
77
+ "sliding_window": null,
78
+ "tie_word_embeddings": true,
79
+ "use_cache": true,
80
+ "use_mrope": false,
81
+ "use_sliding_window": false,
82
+ "vocab_size": 151936
83
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/qwen2/Qwen/Qwen2.5-0.5B/b558da325694139bbaaa.json ADDED
@@ -0,0 +1,83 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "Qwen/Qwen2.5-0.5B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Qwen2ForCausalLM"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "dtype": "bfloat16",
10
+ "hidden_act": "silu",
11
+ "hidden_size": 896,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 4864,
14
+ "layer_types": [
15
+ "full_attention",
16
+ "full_attention",
17
+ "full_attention",
18
+ "full_attention",
19
+ "full_attention",
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention"
39
+ ],
40
+ "max_position_embeddings": 32768,
41
+ "max_window_layers": 24,
42
+ "model_type": "qwen2",
43
+ "neuron": {
44
+ "_serialized_key": "NxDNeuronConfig",
45
+ "batch_size": 1,
46
+ "capacity_factor": null,
47
+ "checkpoint_id": "Qwen/Qwen2.5-0.5B",
48
+ "checkpoint_revision": "060db6499f32faf8b98477b0a26969ef7d8b9987",
49
+ "continuous_batching": false,
50
+ "ep_degree": 1,
51
+ "fused_qkv": false,
52
+ "glu_mlp": true,
53
+ "local_ranks_size": 2,
54
+ "max_batch_size": 1,
55
+ "max_context_length": 8192,
56
+ "max_topk": 256,
57
+ "n_active_tokens": 8192,
58
+ "neuronxcc_version": "2.21.33363.0+82129205",
59
+ "on_device_sampling": true,
60
+ "optimum_neuron_version": "0.4.4.dev2",
61
+ "output_logits": false,
62
+ "pp_degree": 1,
63
+ "sequence_length": 8192,
64
+ "sequence_parallel_enabled": false,
65
+ "speculation_length": 0,
66
+ "start_rank_id": 0,
67
+ "target": "trn1",
68
+ "torch_dtype": "bfloat16",
69
+ "tp_degree": 2
70
+ },
71
+ "num_attention_heads": 14,
72
+ "num_hidden_layers": 24,
73
+ "num_key_value_heads": 2,
74
+ "rms_norm_eps": 1e-06,
75
+ "rope_scaling": null,
76
+ "rope_theta": 1000000.0,
77
+ "sliding_window": null,
78
+ "tie_word_embeddings": true,
79
+ "use_cache": true,
80
+ "use_mrope": false,
81
+ "use_sliding_window": false,
82
+ "vocab_size": 151936
83
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/qwen3/Qwen/Qwen3-0.6B/1b21de7b61f3d14bbbff.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": 2,
60
+ "max_batch_size": 4,
61
+ "max_context_length": 4096,
62
+ "max_topk": 256,
63
+ "n_active_tokens": 4096,
64
+ "neuronxcc_version": "2.21.33363.0+82129205",
65
+ "on_device_sampling": false,
66
+ "optimum_neuron_version": "0.4.4.dev2",
67
+ "output_logits": false,
68
+ "pp_degree": 1,
69
+ "sequence_length": 4096,
70
+ "sequence_parallel_enabled": false,
71
+ "speculation_length": 0,
72
+ "start_rank_id": 0,
73
+ "target": "trn1",
74
+ "torch_dtype": "bfloat16",
75
+ "tp_degree": 2
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-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/qwen3/Qwen/Qwen3-0.6B/ad8740a6bddc8f6b87a7.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": 1,
52
+ "capacity_factor": null,
53
+ "checkpoint_id": "Qwen/Qwen3-0.6B",
54
+ "checkpoint_revision": "c1899de289a04d12100db370d81485cdf75e47ca",
55
+ "continuous_batching": false,
56
+ "ep_degree": 1,
57
+ "fused_qkv": true,
58
+ "glu_mlp": true,
59
+ "local_ranks_size": 2,
60
+ "max_batch_size": 1,
61
+ "max_context_length": 8192,
62
+ "max_topk": 256,
63
+ "n_active_tokens": 8192,
64
+ "neuronxcc_version": "2.21.33363.0+82129205",
65
+ "on_device_sampling": true,
66
+ "optimum_neuron_version": "0.4.4.dev2",
67
+ "output_logits": false,
68
+ "pp_degree": 1,
69
+ "sequence_length": 8192,
70
+ "sequence_parallel_enabled": false,
71
+ "speculation_length": 0,
72
+ "start_rank_id": 0,
73
+ "target": "trn1",
74
+ "torch_dtype": "bfloat16",
75
+ "tp_degree": 2
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-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/smollm3/HuggingFaceTB/SmolLM3-3B/0165b063d3dfd73ec9af.json ADDED
@@ -0,0 +1,135 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "HuggingFaceTB/SmolLM3-3B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "SmolLM3ForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "dtype": "bfloat16",
11
+ "hidden_act": "silu",
12
+ "hidden_size": 2048,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 11008,
15
+ "layer_types": [
16
+ "full_attention",
17
+ "full_attention",
18
+ "full_attention",
19
+ "full_attention",
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention",
45
+ "full_attention",
46
+ "full_attention",
47
+ "full_attention",
48
+ "full_attention",
49
+ "full_attention",
50
+ "full_attention",
51
+ "full_attention"
52
+ ],
53
+ "max_position_embeddings": 65536,
54
+ "max_window_layers": 28,
55
+ "mlp_bias": false,
56
+ "model_type": "smollm3",
57
+ "neuron": {
58
+ "_serialized_key": "NxDNeuronConfig",
59
+ "batch_size": 4,
60
+ "capacity_factor": null,
61
+ "checkpoint_id": "HuggingFaceTB/SmolLM3-3B",
62
+ "checkpoint_revision": "a07cc9a04f16550a088caea529712d1d335b0ac1",
63
+ "continuous_batching": true,
64
+ "ep_degree": 1,
65
+ "fused_qkv": true,
66
+ "glu_mlp": true,
67
+ "local_ranks_size": 2,
68
+ "max_batch_size": 4,
69
+ "max_context_length": 4096,
70
+ "max_topk": 256,
71
+ "n_active_tokens": 4096,
72
+ "neuronxcc_version": "2.21.33363.0+82129205",
73
+ "on_device_sampling": true,
74
+ "optimum_neuron_version": "0.4.4.dev2",
75
+ "output_logits": false,
76
+ "pp_degree": 1,
77
+ "sequence_length": 4096,
78
+ "sequence_parallel_enabled": false,
79
+ "speculation_length": 0,
80
+ "start_rank_id": 0,
81
+ "target": "trn1",
82
+ "torch_dtype": "bfloat16",
83
+ "tp_degree": 2
84
+ },
85
+ "no_rope_layer_interval": 4,
86
+ "no_rope_layers": [
87
+ 1,
88
+ 1,
89
+ 1,
90
+ 0,
91
+ 1,
92
+ 1,
93
+ 1,
94
+ 0,
95
+ 1,
96
+ 1,
97
+ 1,
98
+ 0,
99
+ 1,
100
+ 1,
101
+ 1,
102
+ 0,
103
+ 1,
104
+ 1,
105
+ 1,
106
+ 0,
107
+ 1,
108
+ 1,
109
+ 1,
110
+ 0,
111
+ 1,
112
+ 1,
113
+ 1,
114
+ 0,
115
+ 1,
116
+ 1,
117
+ 1,
118
+ 0,
119
+ 1,
120
+ 1,
121
+ 1,
122
+ 0
123
+ ],
124
+ "num_attention_heads": 16,
125
+ "num_hidden_layers": 36,
126
+ "num_key_value_heads": 4,
127
+ "pretraining_tp": 2,
128
+ "rms_norm_eps": 1e-06,
129
+ "rope_scaling": null,
130
+ "rope_theta": 5000000.0,
131
+ "sliding_window": null,
132
+ "use_cache": false,
133
+ "use_sliding_window": false,
134
+ "vocab_size": 128256
135
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/smollm3/HuggingFaceTB/SmolLM3-3B/71933d2c6e6099113f93.json ADDED
@@ -0,0 +1,135 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "HuggingFaceTB/SmolLM3-3B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "SmolLM3ForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "dtype": "bfloat16",
11
+ "hidden_act": "silu",
12
+ "hidden_size": 2048,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 11008,
15
+ "layer_types": [
16
+ "full_attention",
17
+ "full_attention",
18
+ "full_attention",
19
+ "full_attention",
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention",
45
+ "full_attention",
46
+ "full_attention",
47
+ "full_attention",
48
+ "full_attention",
49
+ "full_attention",
50
+ "full_attention",
51
+ "full_attention"
52
+ ],
53
+ "max_position_embeddings": 65536,
54
+ "max_window_layers": 28,
55
+ "mlp_bias": false,
56
+ "model_type": "smollm3",
57
+ "neuron": {
58
+ "_serialized_key": "NxDNeuronConfig",
59
+ "batch_size": 1,
60
+ "capacity_factor": null,
61
+ "checkpoint_id": "HuggingFaceTB/SmolLM3-3B",
62
+ "checkpoint_revision": "a07cc9a04f16550a088caea529712d1d335b0ac1",
63
+ "continuous_batching": false,
64
+ "ep_degree": 1,
65
+ "fused_qkv": true,
66
+ "glu_mlp": true,
67
+ "local_ranks_size": 2,
68
+ "max_batch_size": 1,
69
+ "max_context_length": 8192,
70
+ "max_topk": 256,
71
+ "n_active_tokens": 8192,
72
+ "neuronxcc_version": "2.21.33363.0+82129205",
73
+ "on_device_sampling": true,
74
+ "optimum_neuron_version": "0.4.4.dev2",
75
+ "output_logits": false,
76
+ "pp_degree": 1,
77
+ "sequence_length": 8192,
78
+ "sequence_parallel_enabled": true,
79
+ "speculation_length": 0,
80
+ "start_rank_id": 0,
81
+ "target": "trn1",
82
+ "torch_dtype": "bfloat16",
83
+ "tp_degree": 2
84
+ },
85
+ "no_rope_layer_interval": 4,
86
+ "no_rope_layers": [
87
+ 1,
88
+ 1,
89
+ 1,
90
+ 0,
91
+ 1,
92
+ 1,
93
+ 1,
94
+ 0,
95
+ 1,
96
+ 1,
97
+ 1,
98
+ 0,
99
+ 1,
100
+ 1,
101
+ 1,
102
+ 0,
103
+ 1,
104
+ 1,
105
+ 1,
106
+ 0,
107
+ 1,
108
+ 1,
109
+ 1,
110
+ 0,
111
+ 1,
112
+ 1,
113
+ 1,
114
+ 0,
115
+ 1,
116
+ 1,
117
+ 1,
118
+ 0,
119
+ 1,
120
+ 1,
121
+ 1,
122
+ 0
123
+ ],
124
+ "num_attention_heads": 16,
125
+ "num_hidden_layers": 36,
126
+ "num_key_value_heads": 4,
127
+ "pretraining_tp": 2,
128
+ "rms_norm_eps": 1e-06,
129
+ "rope_scaling": null,
130
+ "rope_theta": 5000000.0,
131
+ "sliding_window": null,
132
+ "use_cache": false,
133
+ "use_sliding_window": false,
134
+ "vocab_size": 128256
135
+ }
neuronxcc-2.21.33363.0+82129205/MODULE_110fc80e89006393f738+24129607/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=1", "--logfile=/tmp/nxd_model/context_encoding/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.21.33363.0+82129205/MODULE_110fc80e89006393f738+24129607/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_110fc80e89006393f738+24129607/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60325dbe4e341d0836f3b68cedc6beaa300459af2c947445da4a545a3f703c10
3
+ size 939272
neuronxcc-2.21.33363.0+82129205/MODULE_110fc80e89006393f738+24129607/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e189bba6f572c67c0bd6e9bba05b58d4283f34c50588dd6aa97a34007311ba28
3
+ size 6667264
neuronxcc-2.21.33363.0+82129205/MODULE_196d1b2148ed8629b154+24129607/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=1", "--logfile=/tmp/nxd_model/context_encoding/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.21.33363.0+82129205/MODULE_196d1b2148ed8629b154+24129607/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_196d1b2148ed8629b154+24129607/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d18c81c6ab7ffca4593ffd12280271d86ca9a2700a2770da93fb314a9109ff67
3
+ size 509380
neuronxcc-2.21.33363.0+82129205/MODULE_196d1b2148ed8629b154+24129607/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:45d13b5a6f47ee0e1c7c9b783470687dfcad53929b4035be0b071bbacd718948
3
+ size 41585664
neuronxcc-2.21.33363.0+82129205/MODULE_1aa5f4baa9354745d6a6+24129607/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=1", "--logfile=/tmp/nxd_model/context_encoding/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.21.33363.0+82129205/MODULE_1aa5f4baa9354745d6a6+24129607/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_1aa5f4baa9354745d6a6+24129607/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3ed4ca4ead0b4a889a31f35680291eb4368e838532e40337ccf0a72e56de3aa1
3
+ size 914383
neuronxcc-2.21.33363.0+82129205/MODULE_1aa5f4baa9354745d6a6+24129607/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b4a8ac13c543cfd26597bdef4bb0dc1c5fe6dafe9046db1c62b7d4e1fe1eb4bd
3
+ size 12606464
neuronxcc-2.21.33363.0+82129205/MODULE_37d321becc90cb687039+24129607/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=1", "--logfile=/tmp/nxd_model/context_encoding/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.21.33363.0+82129205/MODULE_37d321becc90cb687039+24129607/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_37d321becc90cb687039+24129607/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b254a5de285d4202654c3861698c1a55558bbae98505f678539066015a5d803a
3
+ size 474402
neuronxcc-2.21.33363.0+82129205/MODULE_37d321becc90cb687039+24129607/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:02b5479a35276997f52e0427e90139634e919e308bc9747f3ca820dec15d0382
3
+ size 115057664
neuronxcc-2.21.33363.0+82129205/MODULE_39061f1efbca2332dc73+24129607/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=1", "--logfile=/tmp/nxd_model/context_encoding/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.21.33363.0+82129205/MODULE_39061f1efbca2332dc73+24129607/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_39061f1efbca2332dc73+24129607/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2bcd5dc785b88fb271b709071dcba90051a1052d0548e3eb0dfd4d88940f24d8
3
+ size 1036312
neuronxcc-2.21.33363.0+82129205/MODULE_39061f1efbca2332dc73+24129607/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:61e6e0d539f585138ff57fe8510d76ba99dc0f679c7886ffa295bb8396c1bd9f
3
+ size 15893504
neuronxcc-2.21.33363.0+82129205/MODULE_5cfe268b844f7d2286a5+a02c3a36/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=1", "--logfile=/tmp/nxd_model/token_generation/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
neuronxcc-2.21.33363.0+82129205/MODULE_5cfe268b844f7d2286a5+a02c3a36/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_5cfe268b844f7d2286a5+a02c3a36/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:92413edeab8bdd1ff79c68aeeb0633faf78e28df1df24ce6e6685975d7156ff1
3
+ size 426769
neuronxcc-2.21.33363.0+82129205/MODULE_5cfe268b844f7d2286a5+a02c3a36/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3c238600da6a768547204d176f9f7adf55db06bcaf6e1cc9a978896b92c42e21
3
+ size 3073024
neuronxcc-2.21.33363.0+82129205/MODULE_5cfe268b844f7d2286a5+a02c3a36/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:54e033c437c96db2869f8b8620b5b2eb24b6fad3e04fd4f0b417f6df1c3558ef
3
+ size 3147125
neuronxcc-2.21.33363.0+82129205/MODULE_7c1ab0225123c184780f+24129607/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=1", "--logfile=/tmp/nxd_model/context_encoding/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.21.33363.0+82129205/MODULE_7c1ab0225123c184780f+24129607/model.done ADDED
File without changes