Hugging Face's logo

optimum-internal-testing
/
neuron-testing-cache

Model card Files
xet
 Community
tengomucho HF Staff commited on
Commit
4837f89
·
verified ·
1 Parent(s): 16b9454

Synchronizing local compiler cache.

Browse files
Files changed (7) hide show
  1. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev2/441269935591cad8d370e512c0b93cdd2fce6247c40e5a4866d872ee5338b0de/8d757b039a90a33816e7.json +81 -0
  2. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev2/gemma3_text/unsloth/gemma-3-270m-it/8d757b039a90a33816e7.json +81 -0
  3. neuronxcc-2.21.33363.0+82129205/MODULE_1d5235b31814f7b0b086+a32116a7/model.hlo_module.pb +1 -1
  4. neuronxcc-2.21.33363.0+82129205/MODULE_1d5235b31814f7b0b086+a32116a7/model.neff +1 -1
  5. neuronxcc-2.21.33363.0+82129205/MODULE_5017b58825859b75d379+ac10809c/model.hlo_module.pb +1 -1
  6. neuronxcc-2.21.33363.0+82129205/MODULE_5017b58825859b75d379+ac10809c/model.neff +1 -1
  7. neuronxcc-2.21.33363.0+82129205/MODULE_5017b58825859b75d379+ac10809c/wrapped_neff.hlo +1 -1
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev2/441269935591cad8d370e512c0b93cdd2fce6247c40e5a4866d872ee5338b0de/8d757b039a90a33816e7.json ADDED
@@ -0,0 +1,81 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "unsloth/gemma-3-270m-it",
4
+ "_sliding_window_pattern": 6,
5
+ "_task": "text-generation",
6
+ "architectures": [
7
+ "Gemma3ForCausalLM"
8
+ ],
9
+ "attention_bias": false,
10
+ "attention_dropout": 0.0,
11
+ "attn_logit_softcapping": null,
12
+ "dtype": "bfloat16",
13
+ "final_logit_softcapping": null,
14
+ "head_dim": 256,
15
+ "hidden_activation": "gelu_pytorch_tanh",
16
+ "hidden_size": 640,
17
+ "initializer_range": 0.02,
18
+ "intermediate_size": 2048,
19
+ "layer_types": [
20
+ "sliding_attention",
21
+ "sliding_attention",
22
+ "sliding_attention",
23
+ "sliding_attention",
24
+ "sliding_attention",
25
+ "full_attention",
26
+ "sliding_attention",
27
+ "sliding_attention",
28
+ "sliding_attention",
29
+ "sliding_attention",
30
+ "sliding_attention",
31
+ "full_attention",
32
+ "sliding_attention",
33
+ "sliding_attention",
34
+ "sliding_attention",
35
+ "sliding_attention",
36
+ "sliding_attention",
37
+ "full_attention"
38
+ ],
39
+ "max_position_embeddings": 32768,
40
+ "model_type": "gemma3_text",
41
+ "neuron": {
42
+ "_serialized_key": "NxDNeuronConfig",
43
+ "batch_size": 1,
44
+ "capacity_factor": null,
45
+ "checkpoint_id": "unsloth/gemma-3-270m-it",
46
+ "checkpoint_revision": "23cf460f6bb16954176b3ddcc8d4f250501458a9",
47
+ "continuous_batching": false,
48
+ "ep_degree": 1,
49
+ "fused_qkv": true,
50
+ "glu_mlp": true,
51
+ "local_ranks_size": 4,
52
+ "max_batch_size": 1,
53
+ "max_context_length": 8192,
54
+ "max_topk": 256,
55
+ "n_active_tokens": 8192,
56
+ "neuronxcc_version": "2.21.33363.0+82129205",
57
+ "on_device_sampling": true,
58
+ "optimum_neuron_version": "0.4.5.dev2",
59
+ "output_logits": false,
60
+ "pp_degree": 1,
61
+ "sequence_length": 8192,
62
+ "speculation_length": 0,
63
+ "start_rank_id": 0,
64
+ "target": "trn2",
65
+ "torch_dtype": "bfloat16",
66
+ "tp_degree": 4
67
+ },
68
+ "num_attention_heads": 4,
69
+ "num_hidden_layers": 18,
70
+ "num_key_value_heads": 1,
71
+ "query_pre_attn_scalar": 256,
72
+ "rms_norm_eps": 1e-06,
73
+ "rope_local_base_freq": 10000.0,
74
+ "rope_scaling": null,
75
+ "rope_theta": 1000000.0,
76
+ "sliding_window": 512,
77
+ "unsloth_fixed": true,
78
+ "use_bidirectional_attention": false,
79
+ "use_cache": true,
80
+ "vocab_size": 262144
81
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev2/gemma3_text/unsloth/gemma-3-270m-it/8d757b039a90a33816e7.json ADDED
@@ -0,0 +1,81 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "unsloth/gemma-3-270m-it",
4
+ "_sliding_window_pattern": 6,
5
+ "_task": "text-generation",
6
+ "architectures": [
7
+ "Gemma3ForCausalLM"
8
+ ],
9
+ "attention_bias": false,
10
+ "attention_dropout": 0.0,
11
+ "attn_logit_softcapping": null,
12
+ "dtype": "bfloat16",
13
+ "final_logit_softcapping": null,
14
+ "head_dim": 256,
15
+ "hidden_activation": "gelu_pytorch_tanh",
16
+ "hidden_size": 640,
17
+ "initializer_range": 0.02,
18
+ "intermediate_size": 2048,
19
+ "layer_types": [
20
+ "sliding_attention",
21
+ "sliding_attention",
22
+ "sliding_attention",
23
+ "sliding_attention",
24
+ "sliding_attention",
25
+ "full_attention",
26
+ "sliding_attention",
27
+ "sliding_attention",
28
+ "sliding_attention",
29
+ "sliding_attention",
30
+ "sliding_attention",
31
+ "full_attention",
32
+ "sliding_attention",
33
+ "sliding_attention",
34
+ "sliding_attention",
35
+ "sliding_attention",
36
+ "sliding_attention",
37
+ "full_attention"
38
+ ],
39
+ "max_position_embeddings": 32768,
40
+ "model_type": "gemma3_text",
41
+ "neuron": {
42
+ "_serialized_key": "NxDNeuronConfig",
43
+ "batch_size": 1,
44
+ "capacity_factor": null,
45
+ "checkpoint_id": "unsloth/gemma-3-270m-it",
46
+ "checkpoint_revision": "23cf460f6bb16954176b3ddcc8d4f250501458a9",
47
+ "continuous_batching": false,
48
+ "ep_degree": 1,
49
+ "fused_qkv": true,
50
+ "glu_mlp": true,
51
+ "local_ranks_size": 4,
52
+ "max_batch_size": 1,
53
+ "max_context_length": 8192,
54
+ "max_topk": 256,
55
+ "n_active_tokens": 8192,
56
+ "neuronxcc_version": "2.21.33363.0+82129205",
57
+ "on_device_sampling": true,
58
+ "optimum_neuron_version": "0.4.5.dev2",
59
+ "output_logits": false,
60
+ "pp_degree": 1,
61
+ "sequence_length": 8192,
62
+ "speculation_length": 0,
63
+ "start_rank_id": 0,
64
+ "target": "trn2",
65
+ "torch_dtype": "bfloat16",
66
+ "tp_degree": 4
67
+ },
68
+ "num_attention_heads": 4,
69
+ "num_hidden_layers": 18,
70
+ "num_key_value_heads": 1,
71
+ "query_pre_attn_scalar": 256,
72
+ "rms_norm_eps": 1e-06,
73
+ "rope_local_base_freq": 10000.0,
74
+ "rope_scaling": null,
75
+ "rope_theta": 1000000.0,
76
+ "sliding_window": 512,
77
+ "unsloth_fixed": true,
78
+ "use_bidirectional_attention": false,
79
+ "use_cache": true,
80
+ "vocab_size": 262144
81
+ }
neuronxcc-2.21.33363.0+82129205/MODULE_1d5235b31814f7b0b086+a32116a7/model.hlo_module.pb CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:76a1721d083e2fc227a2ee203458da229a997bde3a0357debd82d3b717f13975
3
  size 1704954
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fcc72101146bb93b226e94d352e88db2d1725fcb2a9c431cf17f35e86718f555
3
  size 1704954
neuronxcc-2.21.33363.0+82129205/MODULE_1d5235b31814f7b0b086+a32116a7/model.neff CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b339508632effabd71cd4866e8c5689d408a7a9fc5dcb45c97399f87509b1256
3
  size 17183744
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9eaca1c813cd8b22f99d2aa46c641493affc50f5d42b82090d885e5d3816dfbc
3
  size 17183744
neuronxcc-2.21.33363.0+82129205/MODULE_5017b58825859b75d379+ac10809c/model.hlo_module.pb CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:74325f682cdf4ccb633a2d974758899a6b732a130f48f66853187ffb02c76124
3
  size 1716991
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2af7df2ce11df8830cbb3866b5c793e0d903d22b41d5155b67a40afe38fd935b
3
  size 1716991
neuronxcc-2.21.33363.0+82129205/MODULE_5017b58825859b75d379+ac10809c/model.neff CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:03e04caf88266ac9bf1cd09383c4b6815ad6b68d5633eec33747654e8acf1f9c
3
  size 1629184
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:230fd54d965dfc4e5d8884f993a011c81452089b354a760a0105b845d16c055e
3
  size 1629184
neuronxcc-2.21.33363.0+82129205/MODULE_5017b58825859b75d379+ac10809c/wrapped_neff.hlo CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ca5f78816ca0d59c1869c148c63133f4c850a4a30a1b9fdb2f04ba63e7dcab05
3
  size 1751074
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:88cd7128c441a75586b7b6c1b826e89a91406f4290765352c1d41405a0e38fe9
3
  size 1751074