Hugging Face's logo

aws-neuron
/
optimum-neuron-cache

Model card Files
xet
Community
586
dacorvo HF Staff commited on
Commit
595454d
·
verified ·
1 Parent(s): 728744c

Synchronizing local compiler cache.

Browse files
Files changed (17) hide show
  1. .gitattributes +4 -0
  2. neuronxcc-2.17.194.0+d312836f/0_REGISTRY/0.3.0.dev1/qwen2/Qwen/Qwen2.5-7B-Instruct/4b31d55cb498352e888f.json +70 -0
  3. neuronxcc-2.17.194.0+d312836f/MODULE_006812725439693f05c7+165e9558/compile_flags.json +1 -0
  4. neuronxcc-2.17.194.0+d312836f/MODULE_006812725439693f05c7+165e9558/model.done +0 -0
  5. neuronxcc-2.17.194.0+d312836f/MODULE_006812725439693f05c7+165e9558/model.hlo_module.pb +3 -0
  6. neuronxcc-2.17.194.0+d312836f/MODULE_006812725439693f05c7+165e9558/model.neff +3 -0
  7. neuronxcc-2.17.194.0+d312836f/MODULE_006812725439693f05c7+165e9558/wrapped_neff.hlo +3 -0
  8. neuronxcc-2.17.194.0+d312836f/MODULE_2541f3cc052802e8251f+bfe5714b/compile_flags.json +1 -0
  9. neuronxcc-2.17.194.0+d312836f/MODULE_2541f3cc052802e8251f+bfe5714b/model.done +0 -0
  10. neuronxcc-2.17.194.0+d312836f/MODULE_2541f3cc052802e8251f+bfe5714b/model.hlo_module.pb +3 -0
  11. neuronxcc-2.17.194.0+d312836f/MODULE_2541f3cc052802e8251f+bfe5714b/model.neff +3 -0
  12. neuronxcc-2.17.194.0+d312836f/MODULE_5474bb5858682409015d+165e9558/compile_flags.json +1 -0
  13. neuronxcc-2.17.194.0+d312836f/MODULE_5474bb5858682409015d+165e9558/model.hlo_module.pb +3 -0
  14. neuronxcc-2.17.194.0+d312836f/MODULE_e30c9836c45db850c730+431f5505/compile_flags.json +1 -0
  15. neuronxcc-2.17.194.0+d312836f/MODULE_e30c9836c45db850c730+431f5505/model.done +0 -0
  16. neuronxcc-2.17.194.0+d312836f/MODULE_e30c9836c45db850c730+431f5505/model.hlo_module.pb +3 -0
  17. neuronxcc-2.17.194.0+d312836f/MODULE_e30c9836c45db850c730+431f5505/model.neff +3 -0
.gitattributes CHANGED
@@ -8527,3 +8527,7 @@ neuronxcc-2.17.194.0+d312836f/MODULE_ddd8df8ccbf7e2239d19+bfe5714b/model.neff fi
8527
  neuronxcc-2.17.194.0+d312836f/MODULE_664bf219018e6f974af9+bfe5714b/model.neff filter=lfs diff=lfs merge=lfs -text
8528
  neuronxcc-2.17.194.0+d312836f/MODULE_accabfa5cf855304ea18+165e9558/model.neff filter=lfs diff=lfs merge=lfs -text
8529
  neuronxcc-2.17.194.0+d312836f/MODULE_accabfa5cf855304ea18+165e9558/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
8527
  neuronxcc-2.17.194.0+d312836f/MODULE_664bf219018e6f974af9+bfe5714b/model.neff filter=lfs diff=lfs merge=lfs -text
8528
  neuronxcc-2.17.194.0+d312836f/MODULE_accabfa5cf855304ea18+165e9558/model.neff filter=lfs diff=lfs merge=lfs -text
8529
  neuronxcc-2.17.194.0+d312836f/MODULE_accabfa5cf855304ea18+165e9558/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
8530
+ neuronxcc-2.17.194.0+d312836f/MODULE_006812725439693f05c7+165e9558/model.neff filter=lfs diff=lfs merge=lfs -text
8531
+ neuronxcc-2.17.194.0+d312836f/MODULE_006812725439693f05c7+165e9558/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
8532
+ neuronxcc-2.17.194.0+d312836f/MODULE_2541f3cc052802e8251f+bfe5714b/model.neff filter=lfs diff=lfs merge=lfs -text
8533
+ neuronxcc-2.17.194.0+d312836f/MODULE_e30c9836c45db850c730+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.17.194.0+d312836f/0_REGISTRY/0.3.0.dev1/qwen2/Qwen/Qwen2.5-7B-Instruct/4b31d55cb498352e888f.json ADDED
@@ -0,0 +1,70 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "Qwen/Qwen2.5-7B-Instruct",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Qwen2ForCausalLM"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "hidden_act": "silu",
10
+ "hidden_size": 3584,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 18944,
13
+ "max_position_embeddings": 32768,
14
+ "max_window_layers": 28,
15
+ "model_type": "qwen2",
16
+ "neuron": {
17
+ "_serialized_key": "NxDNeuronConfig",
18
+ "async_mode": false,
19
+ "attn_kernel_enabled": false,
20
+ "batch_size": 1,
21
+ "capacity_factor": null,
22
+ "cc_pipeline_tiling_factor": 2,
23
+ "checkpoint_id": "Qwen/Qwen2.5-7B-Instruct",
24
+ "checkpoint_revision": "a09a35458c702b33eeacc393d103063234e8bc28",
25
+ "continuous_batching": false,
26
+ "enable_bucketing": false,
27
+ "ep_degree": 1,
28
+ "flash_decoding_enabled": false,
29
+ "fused_qkv": false,
30
+ "glu_mlp": true,
31
+ "is_chunked_prefill": false,
32
+ "local_ranks_size": 8,
33
+ "logical_nc_config": 1,
34
+ "max_batch_size": 1,
35
+ "max_context_length": 4096,
36
+ "max_topk": 256,
37
+ "mlp_kernel_enabled": false,
38
+ "mlp_kernel_fuse_residual_add": false,
39
+ "n_active_tokens": 4096,
40
+ "neuronxcc_version": "2.17.194.0+d312836f",
41
+ "num_cores_per_group": 1,
42
+ "on_device_sampling": true,
43
+ "optimum_neuron_version": "0.3.0.dev1",
44
+ "output_logits": false,
45
+ "padding_side": "right",
46
+ "pp_degree": 1,
47
+ "qk_layernorm": false,
48
+ "qkv_kernel_enabled": false,
49
+ "rpl_reduce_dtype": "bfloat16",
50
+ "sequence_length": 4096,
51
+ "sequence_parallel_enabled": false,
52
+ "speculation_length": 0,
53
+ "start_rank_id": 0,
54
+ "target": null,
55
+ "torch_dtype": "bfloat16",
56
+ "tp_degree": 8,
57
+ "vocab_parallel": false
58
+ },
59
+ "num_attention_heads": 28,
60
+ "num_hidden_layers": 28,
61
+ "num_key_value_heads": 4,
62
+ "rms_norm_eps": 1e-06,
63
+ "rope_scaling": null,
64
+ "rope_theta": 1000000.0,
65
+ "sliding_window": 131072,
66
+ "tie_word_embeddings": false,
67
+ "use_cache": true,
68
+ "use_sliding_window": false,
69
+ "vocab_size": 152064
70
+ }
neuronxcc-2.17.194.0+d312836f/MODULE_006812725439693f05c7+165e9558/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 ' -O2 --internal-num-neuroncores-per-sengine=1 --logfile=/tmp/nxd_model/token_generation_model/_tp0_bk0/log-neuron-cc.txt --enable-internal-neff-wrapper"
neuronxcc-2.17.194.0+d312836f/MODULE_006812725439693f05c7+165e9558/model.done ADDED
File without changes
neuronxcc-2.17.194.0+d312836f/MODULE_006812725439693f05c7+165e9558/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7da7a3e3892b5b1e1afd0e46ec4fda5be2640f62424bd9da3171389558682c1c
3
+ size 717474
neuronxcc-2.17.194.0+d312836f/MODULE_006812725439693f05c7+165e9558/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:52a13eb3ffde0ed704f7a539705e2429b0df7bddae3adcb9db68f0ff641f13d2
3
+ size 2233344
neuronxcc-2.17.194.0+d312836f/MODULE_006812725439693f05c7+165e9558/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3eb999a8648f9b446647000abd833686d7ad8e77e95719a867f2535d417d13db
3
+ size 2402801
neuronxcc-2.17.194.0+d312836f/MODULE_2541f3cc052802e8251f+bfe5714b/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 ' -O2 --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_2541f3cc052802e8251f+bfe5714b/model.done ADDED
File without changes
neuronxcc-2.17.194.0+d312836f/MODULE_2541f3cc052802e8251f+bfe5714b/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:689fef8c1a13a0e06982e5e1f8a41cafa32b0587ad7f2b58d793d44cef284c78
3
+ size 799430
neuronxcc-2.17.194.0+d312836f/MODULE_2541f3cc052802e8251f+bfe5714b/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5ac6758c45293a059bbdcc71f49dee8efcc54decacc0229dd9e1d25936b5f225
3
+ size 29563904
neuronxcc-2.17.194.0+d312836f/MODULE_5474bb5858682409015d+165e9558/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 ' -O2 --internal-num-neuroncores-per-sengine=1 --logfile=/tmp/nxd_model/token_generation_model/_tp0_bk0/log-neuron-cc.txt --enable-internal-neff-wrapper"
neuronxcc-2.17.194.0+d312836f/MODULE_5474bb5858682409015d+165e9558/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7e4f2c669267d15a55eda3bfe68e5d756be7242365610e72858d7d72a744fcc4
3
+ size 699089
neuronxcc-2.17.194.0+d312836f/MODULE_e30c9836c45db850c730+431f5505/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ "--model-type=transformer -O1 --lnc=1 --internal-hlo2tensorizer-options=--experimental-unsafe-fp8e4m3fn-as-fp8e4m3 --logfile=/tmp/nxd_model/layout_opt/log-neuron-cc.txt"
neuronxcc-2.17.194.0+d312836f/MODULE_e30c9836c45db850c730+431f5505/model.done ADDED
File without changes
neuronxcc-2.17.194.0+d312836f/MODULE_e30c9836c45db850c730+431f5505/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b895616d5f0ece5f366c1eb18ae799ba9c055d52ed47227217cd98952322e71d
3
+ size 174781
neuronxcc-2.17.194.0+d312836f/MODULE_e30c9836c45db850c730+431f5505/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d2bdf206f23887bb2ee129c54f57c69946b8b4e3afa09e93c987fef483010927
3
+ size 1721344