optimum-internal-testing-user commited on
Commit
5f168eb
·
verified ·
1 Parent(s): 7d77adb

Synchronizing local compiler cache.

Browse files
.gitattributes CHANGED
@@ -3152,3 +3152,7 @@ neuronxcc-2.18.121.0+9e31e41a/MODULE_a08fd8736783f45e5392+84f3e719/model.neff fi
3152
  neuronxcc-2.18.121.0+9e31e41a/MODULE_e5f864ff8f5ba3001616+5be477de/model.neff filter=lfs diff=lfs merge=lfs -text
3153
  neuronxcc-2.18.121.0+9e31e41a/MODULE_e5f864ff8f5ba3001616+5be477de/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
3154
  neuronxcc-2.18.121.0+9e31e41a/MODULE_f91652bd1c46e6aebbc1+84f3e719/model.neff filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
3152
  neuronxcc-2.18.121.0+9e31e41a/MODULE_e5f864ff8f5ba3001616+5be477de/model.neff filter=lfs diff=lfs merge=lfs -text
3153
  neuronxcc-2.18.121.0+9e31e41a/MODULE_e5f864ff8f5ba3001616+5be477de/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
3154
  neuronxcc-2.18.121.0+9e31e41a/MODULE_f91652bd1c46e6aebbc1+84f3e719/model.neff filter=lfs diff=lfs merge=lfs -text
3155
+ neuronxcc-2.18.121.0+9e31e41a/MODULE_71d6f1724eb9ca6bbf34+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
3156
+ neuronxcc-2.18.121.0+9e31e41a/MODULE_8d4a4e991c9ba3c051b5+5be477de/model.neff filter=lfs diff=lfs merge=lfs -text
3157
+ neuronxcc-2.18.121.0+9e31e41a/MODULE_8d4a4e991c9ba3c051b5+5be477de/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
3158
+ neuronxcc-2.18.121.0+9e31e41a/MODULE_ffcef5c0e4349b0175c7+84f3e719/model.neff filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.18.121.0+9e31e41a/0_REGISTRY/0.3.0.dev1/phi3/microsoft/Phi-3-mini-4k-instruct/9cef5fad50d47db6b597.json ADDED
@@ -0,0 +1,77 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "microsoft/Phi-3-mini-4k-instruct",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Phi3ForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "auto_map": {
11
+ "AutoConfig": "microsoft/Phi-3-mini-4k-instruct--configuration_phi3.Phi3Config",
12
+ "AutoModelForCausalLM": "microsoft/Phi-3-mini-4k-instruct--modeling_phi3.Phi3ForCausalLM"
13
+ },
14
+ "embd_pdrop": 0.0,
15
+ "hidden_act": "silu",
16
+ "hidden_size": 3072,
17
+ "initializer_range": 0.02,
18
+ "intermediate_size": 8192,
19
+ "max_position_embeddings": 4096,
20
+ "model_type": "phi3",
21
+ "neuron": {
22
+ "_serialized_key": "NxDNeuronConfig",
23
+ "async_mode": false,
24
+ "attn_kernel_enabled": false,
25
+ "batch_size": 4,
26
+ "capacity_factor": null,
27
+ "cc_pipeline_tiling_factor": 2,
28
+ "checkpoint_id": "microsoft/Phi-3-mini-4k-instruct",
29
+ "checkpoint_revision": "0a67737cc96d2554230f90338b163bc6380a2a85",
30
+ "continuous_batching": true,
31
+ "enable_bucketing": false,
32
+ "ep_degree": 1,
33
+ "flash_decoding_enabled": false,
34
+ "fused_qkv": true,
35
+ "glu_mlp": true,
36
+ "is_chunked_prefill": false,
37
+ "local_ranks_size": 2,
38
+ "logical_nc_config": 1,
39
+ "max_batch_size": 4,
40
+ "max_context_length": 4096,
41
+ "max_topk": 256,
42
+ "mlp_kernel_enabled": false,
43
+ "mlp_kernel_fuse_residual_add": false,
44
+ "n_active_tokens": 4096,
45
+ "neuronxcc_version": "2.18.121.0+9e31e41a",
46
+ "num_cores_per_group": 1,
47
+ "on_device_sampling": false,
48
+ "optimum_neuron_version": "0.3.0.dev1",
49
+ "output_logits": false,
50
+ "padding_side": "right",
51
+ "pp_degree": 1,
52
+ "qk_layernorm": false,
53
+ "qkv_kernel_enabled": false,
54
+ "rpl_reduce_dtype": "bfloat16",
55
+ "sequence_length": 4096,
56
+ "sequence_parallel_enabled": false,
57
+ "speculation_length": 0,
58
+ "start_rank_id": 0,
59
+ "target": null,
60
+ "torch_dtype": "bfloat16",
61
+ "tp_degree": 2,
62
+ "vocab_parallel": false
63
+ },
64
+ "num_attention_heads": 32,
65
+ "num_hidden_layers": 32,
66
+ "num_key_value_heads": 32,
67
+ "original_max_position_embeddings": 4096,
68
+ "partial_rotary_factor": 1.0,
69
+ "resid_pdrop": 0.0,
70
+ "rms_norm_eps": 1e-05,
71
+ "rope_scaling": null,
72
+ "rope_theta": 10000.0,
73
+ "sliding_window": 2047,
74
+ "tie_word_embeddings": false,
75
+ "use_cache": true,
76
+ "vocab_size": 32064
77
+ }
neuronxcc-2.18.121.0+9e31e41a/MODULE_71d6f1724eb9ca6bbf34+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-2.18.121.0+9e31e41a/MODULE_71d6f1724eb9ca6bbf34+431f5505/model.done ADDED
File without changes
neuronxcc-2.18.121.0+9e31e41a/MODULE_71d6f1724eb9ca6bbf34+431f5505/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3fb7b29b350aa05a5de71272d48a2dda64315a6c16f61d123a3ea9ebb02a0878
3
+ size 136006
neuronxcc-2.18.121.0+9e31e41a/MODULE_71d6f1724eb9ca6bbf34+431f5505/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:93a1bf16131c6a1ebd506e63e809f4c560d4f5495b2a192de056534d358547d4
3
+ size 3974144
neuronxcc-2.18.121.0+9e31e41a/MODULE_8d4a4e991c9ba3c051b5+5be477de/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--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-2.18.121.0+9e31e41a/MODULE_8d4a4e991c9ba3c051b5+5be477de/model.done ADDED
File without changes
neuronxcc-2.18.121.0+9e31e41a/MODULE_8d4a4e991c9ba3c051b5+5be477de/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4cc7eed86b6700ccd46cc05acea01ac27bb63527cba9091e9f815fc4e0fbd232
3
+ size 647962
neuronxcc-2.18.121.0+9e31e41a/MODULE_8d4a4e991c9ba3c051b5+5be477de/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:020706c3236d01a5c380435d76ade6081b34d1a461b90e7e9db6000dc26a4e12
3
+ size 10804224
neuronxcc-2.18.121.0+9e31e41a/MODULE_8d4a4e991c9ba3c051b5+5be477de/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3643e566d47e383a834c50c12954e29d2f76e19804566f261300dddde520e8d2
3
+ size 10941602
neuronxcc-2.18.121.0+9e31e41a/MODULE_ffcef5c0e4349b0175c7+84f3e719/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--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-2.18.121.0+9e31e41a/MODULE_ffcef5c0e4349b0175c7+84f3e719/model.done ADDED
File without changes
neuronxcc-2.18.121.0+9e31e41a/MODULE_ffcef5c0e4349b0175c7+84f3e719/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7bd3a0039ac209ac4ba82aa55159cf2b2936ec8e61258ce7c0f0d3436c020d30
3
+ size 841662
neuronxcc-2.18.121.0+9e31e41a/MODULE_ffcef5c0e4349b0175c7+84f3e719/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:31e67e15a9cc68ff0db908655c62361f47ff0ec35b2d3fbc12158e80e0fb6907
3
+ size 6237184