Hugging Face's logo

optimum-internal-testing
/
neuron-testing-cache

Model card Files Community
dacorvo HF Staff commited on
Commit
1c6589c
·
verified ·
1 Parent(s): 1342061

Synchronizing local compiler cache.

Browse files
Files changed (21) hide show
  1. .gitattributes +4 -0
  2. neuronxcc-2.17.194.0+d312836f/0_REGISTRY/0.2.0.dev5/llama/meta-llama/Meta-Llama-3.1-8B/61ec5ee35df13f5203e3.json +55 -0
  3. neuronxcc-2.17.194.0+d312836f/0_REGISTRY/0.2.0.dev5/llama/unsloth/Llama-3.2-1B-Instruct/bd96975ba59ea098e5c6.json +80 -0
  4. neuronxcc-2.17.194.0+d312836f/MODULE_072764f63c9851e5dac5+613edded/compile_flags.json +1 -0
  5. neuronxcc-2.17.194.0+d312836f/MODULE_072764f63c9851e5dac5+613edded/model.done +0 -0
  6. neuronxcc-2.17.194.0+d312836f/MODULE_072764f63c9851e5dac5+613edded/model.hlo_module.pb +3 -0
  7. neuronxcc-2.17.194.0+d312836f/MODULE_072764f63c9851e5dac5+613edded/model.neff +3 -0
  8. neuronxcc-2.17.194.0+d312836f/MODULE_3010733f7f882761ebd8+613edded/compile_flags.json +1 -0
  9. neuronxcc-2.17.194.0+d312836f/MODULE_3010733f7f882761ebd8+613edded/model.hlo_module.pb +3 -0
  10. neuronxcc-2.17.194.0+d312836f/MODULE_48c6444adcac1b53fda6+613edded/compile_flags.json +1 -0
  11. neuronxcc-2.17.194.0+d312836f/MODULE_48c6444adcac1b53fda6+613edded/model.done +0 -0
  12. neuronxcc-2.17.194.0+d312836f/MODULE_48c6444adcac1b53fda6+613edded/model.hlo_module.pb +3 -0
  13. neuronxcc-2.17.194.0+d312836f/MODULE_48c6444adcac1b53fda6+613edded/model.neff +3 -0
  14. neuronxcc-2.17.194.0+d312836f/MODULE_837e23366798c440ccdc+793f1a96/compile_flags.json +1 -0
  15. neuronxcc-2.17.194.0+d312836f/MODULE_837e23366798c440ccdc+793f1a96/model.done +0 -0
  16. neuronxcc-2.17.194.0+d312836f/MODULE_837e23366798c440ccdc+793f1a96/model.hlo_module.pb +3 -0
  17. neuronxcc-2.17.194.0+d312836f/MODULE_837e23366798c440ccdc+793f1a96/model.neff +3 -0
  18. neuronxcc-2.17.194.0+d312836f/MODULE_cb7223927d0638d20f81+613edded/compile_flags.json +1 -0
  19. neuronxcc-2.17.194.0+d312836f/MODULE_cb7223927d0638d20f81+613edded/model.done +0 -0
  20. neuronxcc-2.17.194.0+d312836f/MODULE_cb7223927d0638d20f81+613edded/model.hlo_module.pb +3 -0
  21. neuronxcc-2.17.194.0+d312836f/MODULE_cb7223927d0638d20f81+613edded/model.neff +3 -0
.gitattributes CHANGED
@@ -2315,3 +2315,7 @@ neuronxcc-2.17.194.0+d312836f/MODULE_f358092e7338b2e4529f+7e4da68b/model.neff fi
2315
  neuronxcc-2.17.194.0+d312836f/MODULE_f358092e7338b2e4529f+7e4da68b/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
2316
  neuronxcc-2.17.194.0+d312836f/MODULE_fa4a1715cab5dbc60401+7e4da68b/model.neff filter=lfs diff=lfs merge=lfs -text
2317
  neuronxcc-2.17.194.0+d312836f/MODULE_fa4a1715cab5dbc60401+7e4da68b/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
2315
  neuronxcc-2.17.194.0+d312836f/MODULE_f358092e7338b2e4529f+7e4da68b/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
2316
  neuronxcc-2.17.194.0+d312836f/MODULE_fa4a1715cab5dbc60401+7e4da68b/model.neff filter=lfs diff=lfs merge=lfs -text
2317
  neuronxcc-2.17.194.0+d312836f/MODULE_fa4a1715cab5dbc60401+7e4da68b/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
2318
+ neuronxcc-2.17.194.0+d312836f/MODULE_072764f63c9851e5dac5+613edded/model.neff filter=lfs diff=lfs merge=lfs -text
2319
+ neuronxcc-2.17.194.0+d312836f/MODULE_48c6444adcac1b53fda6+613edded/model.neff filter=lfs diff=lfs merge=lfs -text
2320
+ neuronxcc-2.17.194.0+d312836f/MODULE_837e23366798c440ccdc+793f1a96/model.neff filter=lfs diff=lfs merge=lfs -text
2321
+ neuronxcc-2.17.194.0+d312836f/MODULE_cb7223927d0638d20f81+613edded/model.neff filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.17.194.0+d312836f/0_REGISTRY/0.2.0.dev5/llama/meta-llama/Meta-Llama-3.1-8B/61ec5ee35df13f5203e3.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": "HloNeuronConfig",
20
+ "all_reduce_dtype": null,
21
+ "allow_flash_attention": true,
22
+ "attention_layout": "BSH",
23
+ "attn_output_transposed": false,
24
+ "auto_cast_type": "bf16",
25
+ "batch_size": 1,
26
+ "checkpoint_id": "meta-llama/Meta-Llama-3.1-8B",
27
+ "checkpoint_revision": "d04e592bb4f6aa9cfee91e2e20afa771667e1d4b",
28
+ "collectives_layout": "HSB",
29
+ "continuous_batching": false,
30
+ "fuse_qkv": true,
31
+ "group_query_attention": "shard-over-heads",
32
+ "log_softmax_scores": false,
33
+ "neuronxcc_version": "2.17.194.0+d312836f",
34
+ "optimum_neuron_version": "0.2.0.dev5",
35
+ "output_all_logits": false,
36
+ "sequence_length": 4096,
37
+ "tp_degree": 8
38
+ },
39
+ "num_attention_heads": 32,
40
+ "num_hidden_layers": 32,
41
+ "num_key_value_heads": 8,
42
+ "pretraining_tp": 1,
43
+ "rms_norm_eps": 1e-05,
44
+ "rope_scaling": {
45
+ "factor": 8.0,
46
+ "high_freq_factor": 4.0,
47
+ "low_freq_factor": 1.0,
48
+ "original_max_position_embeddings": 8192,
49
+ "rope_type": "llama3"
50
+ },
51
+ "rope_theta": 500000.0,
52
+ "tie_word_embeddings": false,
53
+ "use_cache": true,
54
+ "vocab_size": 128256
55
+ }
neuronxcc-2.17.194.0+d312836f/0_REGISTRY/0.2.0.dev5/llama/unsloth/Llama-3.2-1B-Instruct/bd96975ba59ea098e5c6.json ADDED
@@ -0,0 +1,80 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "unsloth/Llama-3.2-1B-Instruct",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "head_dim": 64,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 2048,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 8192,
15
+ "max_position_embeddings": 131072,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "neuron": {
19
+ "_serialized_key": "NxDNeuronConfig",
20
+ "async_mode": false,
21
+ "attn_kernel_enabled": false,
22
+ "batch_size": 4,
23
+ "capacity_factor": null,
24
+ "cc_pipeline_tiling_factor": 2,
25
+ "checkpoint_id": "unsloth/Llama-3.2-1B-Instruct",
26
+ "checkpoint_revision": "5a8abab4a5d6f164389b1079fb721cfab8d7126c",
27
+ "continuous_batching": true,
28
+ "ctx_batch_size": 1,
29
+ "enable_bucketing": false,
30
+ "ep_degree": 1,
31
+ "flash_decoding_enabled": false,
32
+ "fused_qkv": true,
33
+ "glu_mlp": true,
34
+ "is_chunked_prefill": false,
35
+ "local_ranks_size": 2,
36
+ "logical_nc_config": 1,
37
+ "max_batch_size": 4,
38
+ "max_context_length": 4096,
39
+ "max_topk": 256,
40
+ "mlp_kernel_enabled": false,
41
+ "mlp_kernel_fuse_residual_add": false,
42
+ "n_active_tokens": 4096,
43
+ "neuronxcc_version": "2.17.194.0+d312836f",
44
+ "num_cores_per_group": 1,
45
+ "on_device_sampling": true,
46
+ "optimum_neuron_version": "0.2.0.dev5",
47
+ "output_logits": false,
48
+ "padding_side": "right",
49
+ "pp_degree": 1,
50
+ "qk_layernorm": false,
51
+ "qkv_kernel_enabled": false,
52
+ "rpl_reduce_dtype": "float16",
53
+ "sequence_length": 4096,
54
+ "sequence_parallel_enabled": false,
55
+ "speculation_length": 0,
56
+ "start_rank_id": 0,
57
+ "target": null,
58
+ "tkg_batch_size": 4,
59
+ "torch_dtype": "float16",
60
+ "tp_degree": 2,
61
+ "vocab_parallel": false
62
+ },
63
+ "num_attention_heads": 32,
64
+ "num_hidden_layers": 16,
65
+ "num_key_value_heads": 8,
66
+ "pretraining_tp": 1,
67
+ "rms_norm_eps": 1e-05,
68
+ "rope_scaling": {
69
+ "factor": 32.0,
70
+ "high_freq_factor": 4.0,
71
+ "low_freq_factor": 1.0,
72
+ "original_max_position_embeddings": 8192,
73
+ "rope_type": "llama3"
74
+ },
75
+ "rope_theta": 500000.0,
76
+ "tie_word_embeddings": true,
77
+ "unsloth_fixed": true,
78
+ "use_cache": true,
79
+ "vocab_size": 128256
80
+ }
neuronxcc-2.17.194.0+d312836f/MODULE_072764f63c9851e5dac5+613edded/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--model-type=transformer", "--auto-cast=none", "--execute-repetition=1"]
neuronxcc-2.17.194.0+d312836f/MODULE_072764f63c9851e5dac5+613edded/model.done ADDED
File without changes
neuronxcc-2.17.194.0+d312836f/MODULE_072764f63c9851e5dac5+613edded/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cdda8f0916cbf4ea683a5b82d37da6d44e3d8594ffb075276862d47c9784b5c6
3
+ size 421229
neuronxcc-2.17.194.0+d312836f/MODULE_072764f63c9851e5dac5+613edded/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6350f78e621de70a884d9bdd3d20101c72514fc6c874594458c2f6d86e12cdd5
3
+ size 30209024
neuronxcc-2.17.194.0+d312836f/MODULE_3010733f7f882761ebd8+613edded/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--model-type=transformer", "--auto-cast=none", "--execute-repetition=1"]
neuronxcc-2.17.194.0+d312836f/MODULE_3010733f7f882761ebd8+613edded/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:373713a19c37d0213b439380ffb05009fae9869c3a6e00b41dbca565af9b6da5
3
+ size 425388
neuronxcc-2.17.194.0+d312836f/MODULE_48c6444adcac1b53fda6+613edded/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--model-type=transformer", "--auto-cast=none", "--execute-repetition=1"]
neuronxcc-2.17.194.0+d312836f/MODULE_48c6444adcac1b53fda6+613edded/model.done ADDED
File without changes
neuronxcc-2.17.194.0+d312836f/MODULE_48c6444adcac1b53fda6+613edded/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:da33932abafecde481f769f8c365b4389cf32ef68dd270cfeda7fa2ebe556543
3
+ size 331676
neuronxcc-2.17.194.0+d312836f/MODULE_48c6444adcac1b53fda6+613edded/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f5255d986fb2240fb123fe23f9b70b6c0119d53197b40256a564eb5114fbf12c
3
+ size 2571264
neuronxcc-2.17.194.0+d312836f/MODULE_837e23366798c440ccdc+793f1a96/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ "--auto-cast=none --model-type=transformer --tensorizer-options='--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-dge-dma --vectorize-strided-dma ' -O1 --internal-num-neuroncores-per-sengine=1 --logfile=/tmp/nxd_model/context_encoding_model/_tp0_bk0/log-neuron-cc.txt"
neuronxcc-2.17.194.0+d312836f/MODULE_837e23366798c440ccdc+793f1a96/model.done ADDED
File without changes
neuronxcc-2.17.194.0+d312836f/MODULE_837e23366798c440ccdc+793f1a96/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d29567b9d379031562b538d6faf0fbdfe4a0b11782f3bf2e1f3274e541b43bcc
3
+ size 458016
neuronxcc-2.17.194.0+d312836f/MODULE_837e23366798c440ccdc+793f1a96/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6a9c96505aa6bac3f4e2777240679f3ffdb01db4136207b144e8c29c5b4ae32d
3
+ size 6431744
neuronxcc-2.17.194.0+d312836f/MODULE_cb7223927d0638d20f81+613edded/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--model-type=transformer", "--auto-cast=none", "--execute-repetition=1"]
neuronxcc-2.17.194.0+d312836f/MODULE_cb7223927d0638d20f81+613edded/model.done ADDED
File without changes
neuronxcc-2.17.194.0+d312836f/MODULE_cb7223927d0638d20f81+613edded/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5b0c9b97612f7bf7cff942c0feb4c43a02a1d453152afe729985d2f86548c40c
3
+ size 332710
neuronxcc-2.17.194.0+d312836f/MODULE_cb7223927d0638d20f81+613edded/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bde578b178ec4415bdc6c07da2e05ee48e7623dfa22973a71f01b3994b345f5e
3
+ size 4557824