Hugging Face's logo

optimum-internal-testing
/
neuron-testing-cache

Model card Files
xet
Community
dacorvo HF Staff commited on
Commit
4e8fa27
·
verified ·
1 Parent(s): 8a835f4

Synchronizing local compiler cache.

Browse files
Files changed (16) hide show
  1. neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/TinyLlama/TinyLlama-1.1B-Chat-v1.0/60e80fb96b6a062f2a56.json +56 -0
  2. neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Llama-2-13b-hf/a5922a110399492be422.json +56 -0
  3. neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Llama-2-7b-hf/7f22c62ac2872621ab6f.json +56 -0
  4. neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Llama-3.2-1B/a40f79324455596f82b6.json +62 -0
  5. neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Llama-3.2-1B/b8bfbbad0465ef7dfb11.json +62 -0
  6. neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Llama-3.2-3B/144a9a8f0d875b2d168f.json +62 -0
  7. neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Llama-3.2-3B/b484c1ff10d6c31d8bd7.json +62 -0
  8. neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Meta-Llama-3-8B/bb26c6d5320446b556d8.json +56 -0
  9. neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Meta-Llama-3.1-8B/088c9cb62aff91127ee5.json +62 -0
  10. neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Meta-Llama-3.1-8B/6faddb420172ebad7a65.json +62 -0
  11. neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Meta-Llama-3.1-8B/7e972749ba3b6416fefe.json +62 -0
  12. neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Meta-Llama-3.1-8B/93afe7224308bbf65462.json +62 -0
  13. neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Meta-Llama-3.1-8B/9aaca0d9f5b6c41a6516.json +62 -0
  14. neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Meta-Llama-3.1-8B/dd56b3c2b165b1f96cb5.json +62 -0
  15. neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Meta-Llama-3.1-8B/fc43bb8fcc4c0f60544b.json +62 -0
  16. neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.1.dev0/qwen2/Qwen/Qwen2.5-0.5B/7621f6503e2289297f1f.json +82 -0
neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/TinyLlama/TinyLlama-1.1B-Chat-v1.0/60e80fb96b6a062f2a56.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "TinyLlama/TinyLlama-1.1B-Chat-v1.0",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "head_dim": 64,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 2048,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 5632,
15
+ "max_position_embeddings": 2048,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "neuron": {
19
+ "_serialized_key": "NxDNeuronConfig",
20
+ "batch_size": 1,
21
+ "capacity_factor": null,
22
+ "checkpoint_id": "TinyLlama/TinyLlama-1.1B-Chat-v1.0",
23
+ "checkpoint_revision": "fe8a4ea1ffedaf415f4da2f062534de366a451e6",
24
+ "continuous_batching": false,
25
+ "enable_bucketing": 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": 2048,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 2048,
34
+ "neuronxcc_version": "2.21.18209.0+043b1bf7",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.0",
37
+ "output_logits": false,
38
+ "pp_degree": 1,
39
+ "sequence_length": 2048,
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": 22,
48
+ "num_key_value_heads": 4,
49
+ "pretraining_tp": 1,
50
+ "rms_norm_eps": 1e-05,
51
+ "rope_scaling": null,
52
+ "rope_theta": 10000.0,
53
+ "tie_word_embeddings": false,
54
+ "use_cache": true,
55
+ "vocab_size": 32000
56
+ }
neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Llama-2-13b-hf/a5922a110399492be422.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "meta-llama/Llama-2-13b-hf",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 5120,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 13824,
15
+ "max_position_embeddings": 4096,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "neuron": {
19
+ "_serialized_key": "NxDNeuronConfig",
20
+ "batch_size": 1,
21
+ "capacity_factor": null,
22
+ "checkpoint_id": "meta-llama/Llama-2-13b-hf",
23
+ "checkpoint_revision": "5c31dfb671ce7cfe2d7bb7c04375e44c55e815b1",
24
+ "continuous_batching": false,
25
+ "enable_bucketing": false,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 8,
30
+ "max_batch_size": 1,
31
+ "max_context_length": 2048,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 2048,
34
+ "neuronxcc_version": "2.21.18209.0+043b1bf7",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.0",
37
+ "output_logits": false,
38
+ "pp_degree": 1,
39
+ "sequence_length": 2048,
40
+ "speculation_length": 0,
41
+ "start_rank_id": 0,
42
+ "target": "trn1",
43
+ "torch_dtype": "bfloat16",
44
+ "tp_degree": 8
45
+ },
46
+ "num_attention_heads": 40,
47
+ "num_hidden_layers": 40,
48
+ "num_key_value_heads": 40,
49
+ "pretraining_tp": 1,
50
+ "rms_norm_eps": 1e-05,
51
+ "rope_scaling": null,
52
+ "rope_theta": 10000.0,
53
+ "tie_word_embeddings": false,
54
+ "use_cache": true,
55
+ "vocab_size": 32000
56
+ }
neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Llama-2-7b-hf/7f22c62ac2872621ab6f.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "meta-llama/Llama-2-7b-hf",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 4096,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 11008,
15
+ "max_position_embeddings": 4096,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "neuron": {
19
+ "_serialized_key": "NxDNeuronConfig",
20
+ "batch_size": 1,
21
+ "capacity_factor": null,
22
+ "checkpoint_id": "meta-llama/Llama-2-7b-hf",
23
+ "checkpoint_revision": "01c7f73d771dfac7d292323805ebc428287df4f9",
24
+ "continuous_batching": false,
25
+ "enable_bucketing": 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": 2048,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 2048,
34
+ "neuronxcc_version": "2.21.18209.0+043b1bf7",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.0",
37
+ "output_logits": false,
38
+ "pp_degree": 1,
39
+ "sequence_length": 2048,
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": 32,
48
+ "num_key_value_heads": 32,
49
+ "pretraining_tp": 1,
50
+ "rms_norm_eps": 1e-05,
51
+ "rope_scaling": null,
52
+ "rope_theta": 10000.0,
53
+ "tie_word_embeddings": false,
54
+ "use_cache": true,
55
+ "vocab_size": 32000
56
+ }
neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Llama-3.2-1B/a40f79324455596f82b6.json ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "batch_size": 4,
21
+ "capacity_factor": null,
22
+ "checkpoint_id": "meta-llama/Llama-3.2-1B",
23
+ "checkpoint_revision": "4e20de362430cd3b72f300e6b0f18e50e7166e08",
24
+ "continuous_batching": true,
25
+ "enable_bucketing": false,
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": "2.21.18209.0+043b1bf7",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.0",
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
+ "use_cache": true,
61
+ "vocab_size": 128256
62
+ }
neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Llama-3.2-1B/b8bfbbad0465ef7dfb11.json ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "batch_size": 1,
21
+ "capacity_factor": null,
22
+ "checkpoint_id": "meta-llama/Llama-3.2-1B",
23
+ "checkpoint_revision": "4e20de362430cd3b72f300e6b0f18e50e7166e08",
24
+ "continuous_batching": false,
25
+ "enable_bucketing": false,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 2,
30
+ "max_batch_size": 1,
31
+ "max_context_length": 4096,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 4096,
34
+ "neuronxcc_version": "2.21.18209.0+043b1bf7",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.0",
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
+ "use_cache": true,
61
+ "vocab_size": 128256
62
+ }
neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Llama-3.2-3B/144a9a8f0d875b2d168f.json ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "meta-llama/Llama-3.2-3B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 3072,
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
+ "batch_size": 1,
21
+ "capacity_factor": null,
22
+ "checkpoint_id": "meta-llama/Llama-3.2-3B",
23
+ "checkpoint_revision": "13afe5124825b4f3751f836b40dafda64c1ed062",
24
+ "continuous_batching": false,
25
+ "enable_bucketing": false,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 2,
30
+ "max_batch_size": 1,
31
+ "max_context_length": 4096,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 4096,
34
+ "neuronxcc_version": "2.21.18209.0+043b1bf7",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.0",
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": 24,
47
+ "num_hidden_layers": 28,
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
+ "use_cache": true,
61
+ "vocab_size": 128256
62
+ }
neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Llama-3.2-3B/b484c1ff10d6c31d8bd7.json ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "meta-llama/Llama-3.2-3B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 3072,
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
+ "batch_size": 4,
21
+ "capacity_factor": null,
22
+ "checkpoint_id": "meta-llama/Llama-3.2-3B",
23
+ "checkpoint_revision": "13afe5124825b4f3751f836b40dafda64c1ed062",
24
+ "continuous_batching": true,
25
+ "enable_bucketing": false,
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": "2.21.18209.0+043b1bf7",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.0",
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": 24,
47
+ "num_hidden_layers": 28,
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
+ "use_cache": true,
61
+ "vocab_size": 128256
62
+ }
neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Meta-Llama-3-8B/bb26c6d5320446b556d8.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "meta-llama/Meta-Llama-3-8B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 4096,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 14336,
15
+ "max_position_embeddings": 8192,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "neuron": {
19
+ "_serialized_key": "NxDNeuronConfig",
20
+ "batch_size": 1,
21
+ "capacity_factor": null,
22
+ "checkpoint_id": "meta-llama/Meta-Llama-3-8B",
23
+ "checkpoint_revision": "8cde5ca8380496c9a6cc7ef3a8b46a0372a1d920",
24
+ "continuous_batching": false,
25
+ "enable_bucketing": false,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 2,
30
+ "max_batch_size": 1,
31
+ "max_context_length": 4096,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 4096,
34
+ "neuronxcc_version": "2.21.18209.0+043b1bf7",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.0",
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": 32,
48
+ "num_key_value_heads": 8,
49
+ "pretraining_tp": 1,
50
+ "rms_norm_eps": 1e-05,
51
+ "rope_scaling": null,
52
+ "rope_theta": 500000.0,
53
+ "tie_word_embeddings": false,
54
+ "use_cache": true,
55
+ "vocab_size": 128256
56
+ }
neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Meta-Llama-3.1-8B/088c9cb62aff91127ee5.json ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "meta-llama/Meta-Llama-3.1-8B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 4096,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 14336,
15
+ "max_position_embeddings": 131072,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "neuron": {
19
+ "_serialized_key": "NxDNeuronConfig",
20
+ "batch_size": 16,
21
+ "capacity_factor": null,
22
+ "checkpoint_id": "meta-llama/Meta-Llama-3.1-8B",
23
+ "checkpoint_revision": "d04e592bb4f6aa9cfee91e2e20afa771667e1d4b",
24
+ "continuous_batching": true,
25
+ "enable_bucketing": false,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 8,
30
+ "max_batch_size": 16,
31
+ "max_context_length": 4096,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 4096,
34
+ "neuronxcc_version": "2.21.18209.0+043b1bf7",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.0",
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": 8
45
+ },
46
+ "num_attention_heads": 32,
47
+ "num_hidden_layers": 32,
48
+ "num_key_value_heads": 8,
49
+ "pretraining_tp": 1,
50
+ "rms_norm_eps": 1e-05,
51
+ "rope_scaling": {
52
+ "factor": 8.0,
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": false,
60
+ "use_cache": true,
61
+ "vocab_size": 128256
62
+ }
neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Meta-Llama-3.1-8B/6faddb420172ebad7a65.json ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "meta-llama/Meta-Llama-3.1-8B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 4096,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 14336,
15
+ "max_position_embeddings": 131072,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "neuron": {
19
+ "_serialized_key": "NxDNeuronConfig",
20
+ "batch_size": 8,
21
+ "capacity_factor": null,
22
+ "checkpoint_id": "meta-llama/Meta-Llama-3.1-8B",
23
+ "checkpoint_revision": "d04e592bb4f6aa9cfee91e2e20afa771667e1d4b",
24
+ "continuous_batching": true,
25
+ "enable_bucketing": false,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 8,
30
+ "max_batch_size": 8,
31
+ "max_context_length": 4096,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 4096,
34
+ "neuronxcc_version": "2.21.18209.0+043b1bf7",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.0",
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": 8
45
+ },
46
+ "num_attention_heads": 32,
47
+ "num_hidden_layers": 32,
48
+ "num_key_value_heads": 8,
49
+ "pretraining_tp": 1,
50
+ "rms_norm_eps": 1e-05,
51
+ "rope_scaling": {
52
+ "factor": 8.0,
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": false,
60
+ "use_cache": true,
61
+ "vocab_size": 128256
62
+ }
neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Meta-Llama-3.1-8B/7e972749ba3b6416fefe.json ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "meta-llama/Meta-Llama-3.1-8B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 4096,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 14336,
15
+ "max_position_embeddings": 131072,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "neuron": {
19
+ "_serialized_key": "NxDNeuronConfig",
20
+ "batch_size": 8,
21
+ "capacity_factor": null,
22
+ "checkpoint_id": "meta-llama/Meta-Llama-3.1-8B",
23
+ "checkpoint_revision": "d04e592bb4f6aa9cfee91e2e20afa771667e1d4b",
24
+ "continuous_batching": true,
25
+ "enable_bucketing": false,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 2,
30
+ "max_batch_size": 8,
31
+ "max_context_length": 4096,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 4096,
34
+ "neuronxcc_version": "2.21.18209.0+043b1bf7",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.0",
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": 32,
48
+ "num_key_value_heads": 8,
49
+ "pretraining_tp": 1,
50
+ "rms_norm_eps": 1e-05,
51
+ "rope_scaling": {
52
+ "factor": 8.0,
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": false,
60
+ "use_cache": true,
61
+ "vocab_size": 128256
62
+ }
neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Meta-Llama-3.1-8B/93afe7224308bbf65462.json ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "meta-llama/Meta-Llama-3.1-8B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 4096,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 14336,
15
+ "max_position_embeddings": 131072,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "neuron": {
19
+ "_serialized_key": "NxDNeuronConfig",
20
+ "batch_size": 32,
21
+ "capacity_factor": null,
22
+ "checkpoint_id": "meta-llama/Meta-Llama-3.1-8B",
23
+ "checkpoint_revision": "d04e592bb4f6aa9cfee91e2e20afa771667e1d4b",
24
+ "continuous_batching": true,
25
+ "enable_bucketing": false,
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": "2.21.18209.0+043b1bf7",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.0",
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": 8
45
+ },
46
+ "num_attention_heads": 32,
47
+ "num_hidden_layers": 32,
48
+ "num_key_value_heads": 8,
49
+ "pretraining_tp": 1,
50
+ "rms_norm_eps": 1e-05,
51
+ "rope_scaling": {
52
+ "factor": 8.0,
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": false,
60
+ "use_cache": true,
61
+ "vocab_size": 128256
62
+ }
neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Meta-Llama-3.1-8B/9aaca0d9f5b6c41a6516.json ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "meta-llama/Meta-Llama-3.1-8B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 4096,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 14336,
15
+ "max_position_embeddings": 131072,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "neuron": {
19
+ "_serialized_key": "NxDNeuronConfig",
20
+ "batch_size": 1,
21
+ "capacity_factor": null,
22
+ "checkpoint_id": "meta-llama/Meta-Llama-3.1-8B",
23
+ "checkpoint_revision": "d04e592bb4f6aa9cfee91e2e20afa771667e1d4b",
24
+ "continuous_batching": false,
25
+ "enable_bucketing": false,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 2,
30
+ "max_batch_size": 1,
31
+ "max_context_length": 4096,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 4096,
34
+ "neuronxcc_version": "2.21.18209.0+043b1bf7",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.0",
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": 32,
48
+ "num_key_value_heads": 8,
49
+ "pretraining_tp": 1,
50
+ "rms_norm_eps": 1e-05,
51
+ "rope_scaling": {
52
+ "factor": 8.0,
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": false,
60
+ "use_cache": true,
61
+ "vocab_size": 128256
62
+ }
neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Meta-Llama-3.1-8B/dd56b3c2b165b1f96cb5.json ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "meta-llama/Meta-Llama-3.1-8B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 4096,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 14336,
15
+ "max_position_embeddings": 131072,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "neuron": {
19
+ "_serialized_key": "NxDNeuronConfig",
20
+ "batch_size": 4,
21
+ "capacity_factor": null,
22
+ "checkpoint_id": "meta-llama/Meta-Llama-3.1-8B",
23
+ "checkpoint_revision": "d04e592bb4f6aa9cfee91e2e20afa771667e1d4b",
24
+ "continuous_batching": true,
25
+ "enable_bucketing": false,
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": "2.21.18209.0+043b1bf7",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.0",
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": 32,
48
+ "num_key_value_heads": 8,
49
+ "pretraining_tp": 1,
50
+ "rms_norm_eps": 1e-05,
51
+ "rope_scaling": {
52
+ "factor": 8.0,
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": false,
60
+ "use_cache": true,
61
+ "vocab_size": 128256
62
+ }
neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/llama/meta-llama/Meta-Llama-3.1-8B/fc43bb8fcc4c0f60544b.json ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "meta-llama/Meta-Llama-3.1-8B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 4096,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 14336,
15
+ "max_position_embeddings": 131072,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "neuron": {
19
+ "_serialized_key": "NxDNeuronConfig",
20
+ "batch_size": 4,
21
+ "capacity_factor": null,
22
+ "checkpoint_id": "meta-llama/Meta-Llama-3.1-8B",
23
+ "checkpoint_revision": "d04e592bb4f6aa9cfee91e2e20afa771667e1d4b",
24
+ "continuous_batching": true,
25
+ "enable_bucketing": false,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 8,
30
+ "max_batch_size": 4,
31
+ "max_context_length": 4096,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 4096,
34
+ "neuronxcc_version": "2.21.18209.0+043b1bf7",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.0",
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": 8
45
+ },
46
+ "num_attention_heads": 32,
47
+ "num_hidden_layers": 32,
48
+ "num_key_value_heads": 8,
49
+ "pretraining_tp": 1,
50
+ "rms_norm_eps": 1e-05,
51
+ "rope_scaling": {
52
+ "factor": 8.0,
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": false,
60
+ "use_cache": true,
61
+ "vocab_size": 128256
62
+ }
neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.1.dev0/qwen2/Qwen/Qwen2.5-0.5B/7621f6503e2289297f1f.json ADDED
@@ -0,0 +1,82 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "Qwen/Qwen2.5-0.5B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Qwen2ForCausalLM"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "hidden_act": "silu",
10
+ "hidden_size": 896,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 4864,
13
+ "layer_types": [
14
+ "full_attention",
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
+ ],
39
+ "max_position_embeddings": 32768,
40
+ "max_window_layers": 24,
41
+ "model_type": "qwen2",
42
+ "neuron": {
43
+ "_serialized_key": "NxDNeuronConfig",
44
+ "batch_size": 4,
45
+ "capacity_factor": null,
46
+ "checkpoint_id": "Qwen/Qwen2.5-0.5B",
47
+ "checkpoint_revision": "060db6499f32faf8b98477b0a26969ef7d8b9987",
48
+ "continuous_batching": true,
49
+ "enable_bucketing": false,
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": "2.21.18209.0+043b1bf7",
59
+ "on_device_sampling": false,
60
+ "optimum_neuron_version": "0.4.1.dev0",
61
+ "output_logits": false,
62
+ "pp_degree": 1,
63
+ "sequence_length": 4096,
64
+ "speculation_length": 0,
65
+ "start_rank_id": 0,
66
+ "target": "trn1",
67
+ "torch_dtype": "bfloat16",
68
+ "tp_degree": 2
69
+ },
70
+ "num_attention_heads": 14,
71
+ "num_hidden_layers": 24,
72
+ "num_key_value_heads": 2,
73
+ "rms_norm_eps": 1e-06,
74
+ "rope_scaling": null,
75
+ "rope_theta": 1000000.0,
76
+ "sliding_window": null,
77
+ "tie_word_embeddings": true,
78
+ "use_cache": true,
79
+ "use_mrope": false,
80
+ "use_sliding_window": false,
81
+ "vocab_size": 151936
82
+ }