Hugging Face's logo

optimum-internal-testing
/
neuron-testing-cache

Model card Files
xet
 Community
optimum-internal-testing-user commited on
Commit
985a1ad
·
verified ·
1 Parent(s): c3e89d9

Synchronizing local compiler cache.

Browse files
Files changed (7) hide show
  1. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev0/441269935591cad8d370e512c0b93cdd2fce6247c40e5a4866d872ee5338b0de/76d7cb8b92e6056492b5.json +81 -0
  2. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev0/gemma3_text/unsloth/gemma-3-270m-it/76d7cb8b92e6056492b5.json +81 -0
  3. neuronxcc-2.21.33363.0+82129205/MODULE_a155b32467ad91efdb56+a02c3a36/model.hlo_module.pb +1 -1
  4. neuronxcc-2.21.33363.0+82129205/MODULE_a155b32467ad91efdb56+a02c3a36/model.neff +1 -1
  5. neuronxcc-2.21.33363.0+82129205/MODULE_a155b32467ad91efdb56+a02c3a36/wrapped_neff.hlo +1 -1
  6. neuronxcc-2.21.33363.0+82129205/MODULE_e45e2c512692918b8ac7+24129607/model.hlo_module.pb +1 -1
  7. neuronxcc-2.21.33363.0+82129205/MODULE_e45e2c512692918b8ac7+24129607/model.neff +1 -1
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev0/441269935591cad8d370e512c0b93cdd2fce6247c40e5a4866d872ee5338b0de/76d7cb8b92e6056492b5.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": 4,
44
+ "capacity_factor": null,
45
+ "checkpoint_id": "unsloth/gemma-3-270m-it",
46
+ "checkpoint_revision": "23cf460f6bb16954176b3ddcc8d4f250501458a9",
47
+ "continuous_batching": true,
48
+ "ep_degree": 1,
49
+ "fused_qkv": true,
50
+ "glu_mlp": true,
51
+ "local_ranks_size": 2,
52
+ "max_batch_size": 4,
53
+ "max_context_length": 4096,
54
+ "max_topk": 256,
55
+ "n_active_tokens": 4096,
56
+ "neuronxcc_version": "2.21.33363.0+82129205",
57
+ "on_device_sampling": true,
58
+ "optimum_neuron_version": "0.4.6.dev0",
59
+ "output_logits": false,
60
+ "pp_degree": 1,
61
+ "sequence_length": 4096,
62
+ "speculation_length": 0,
63
+ "start_rank_id": 0,
64
+ "target": "trn1",
65
+ "torch_dtype": "bfloat16",
66
+ "tp_degree": 2
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/0.4.6.dev0/gemma3_text/unsloth/gemma-3-270m-it/76d7cb8b92e6056492b5.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": 4,
44
+ "capacity_factor": null,
45
+ "checkpoint_id": "unsloth/gemma-3-270m-it",
46
+ "checkpoint_revision": "23cf460f6bb16954176b3ddcc8d4f250501458a9",
47
+ "continuous_batching": true,
48
+ "ep_degree": 1,
49
+ "fused_qkv": true,
50
+ "glu_mlp": true,
51
+ "local_ranks_size": 2,
52
+ "max_batch_size": 4,
53
+ "max_context_length": 4096,
54
+ "max_topk": 256,
55
+ "n_active_tokens": 4096,
56
+ "neuronxcc_version": "2.21.33363.0+82129205",
57
+ "on_device_sampling": true,
58
+ "optimum_neuron_version": "0.4.6.dev0",
59
+ "output_logits": false,
60
+ "pp_degree": 1,
61
+ "sequence_length": 4096,
62
+ "speculation_length": 0,
63
+ "start_rank_id": 0,
64
+ "target": "trn1",
65
+ "torch_dtype": "bfloat16",
66
+ "tp_degree": 2
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_a155b32467ad91efdb56+a02c3a36/model.hlo_module.pb CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:66f6450490b10388dafb6e4fb250f49f37b11224dd2e5c49f481a0976f729106
3
  size 562167
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cedde36e9447943a3efdf4ed22ba8e8377c18dc164c05dc9ec3b2404fd285d6c
3
  size 562167
neuronxcc-2.21.33363.0+82129205/MODULE_a155b32467ad91efdb56+a02c3a36/model.neff CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b790c945f3cc8a9b7ea96d56aae5fedb2c1a7646ff10fc3036d2a6133674a355
3
  size 1793024
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:26b19d16e87edce0eb519472854464ea428c8c6621ce680463c7a0576186d64b
3
  size 1793024
neuronxcc-2.21.33363.0+82129205/MODULE_a155b32467ad91efdb56+a02c3a36/wrapped_neff.hlo CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c26edc08c497845547cd6794035c39532ad93728ff873571301fd5a96d4d827e
3
  size 1915413
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4364f3e30295be20e1b36a9f8b3877323fbdd44fe0154ef2aa2e7e559a991578
3
  size 1915413
neuronxcc-2.21.33363.0+82129205/MODULE_e45e2c512692918b8ac7+24129607/model.hlo_module.pb CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:709aef187eaab30b99967d754304fe6bf08161b2e4820c8d294053c66f68c8ae
3
  size 571822
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:09002100cbee99f3cf9d355464c3b373418336fa7f26fce3379b075a8ff4dd8b
3
  size 571822
neuronxcc-2.21.33363.0+82129205/MODULE_e45e2c512692918b8ac7+24129607/model.neff CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e24a90f3adcc804a65a44f78af0443d46d2cdb34823f1979b5008dbc53dddbb3
3
  size 9503744
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b94048b1e583bea84cb9e311cfe1f6298e201e617dedcd340fbc12ea64ab735d
3
  size 9503744