optimum-internal-testing-user commited on
Commit
124e6c1
·
verified ·
1 Parent(s): 3802c12

Synchronizing local compiler cache.

Browse files
.gitattributes CHANGED
@@ -3547,3 +3547,7 @@ neuronxcc-2.17.194.0+d312836f/MODULE_1c05973efd8a84bfd97f+165e9558/model.neff fi
3547
  neuronxcc-2.17.194.0+d312836f/MODULE_1c05973efd8a84bfd97f+165e9558/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
3548
  neuronxcc-2.17.194.0+d312836f/MODULE_887e3231a859a1d6272f+bfe5714b/model.neff filter=lfs diff=lfs merge=lfs -text
3549
  neuronxcc-2.17.194.0+d312836f/MODULE_fe5cc26344fc85d3582c+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
3547
  neuronxcc-2.17.194.0+d312836f/MODULE_1c05973efd8a84bfd97f+165e9558/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
3548
  neuronxcc-2.17.194.0+d312836f/MODULE_887e3231a859a1d6272f+bfe5714b/model.neff filter=lfs diff=lfs merge=lfs -text
3549
  neuronxcc-2.17.194.0+d312836f/MODULE_fe5cc26344fc85d3582c+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
3550
+ neuronxcc-2.17.194.0+d312836f/MODULE_04ccf0a18ef2a2487c5a+bfe5714b/model.neff filter=lfs diff=lfs merge=lfs -text
3551
+ neuronxcc-2.17.194.0+d312836f/MODULE_0597e13911f7cd4171ab+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
3552
+ neuronxcc-2.17.194.0+d312836f/MODULE_5f6d188e59db94aef5f0+165e9558/model.neff filter=lfs diff=lfs merge=lfs -text
3553
+ neuronxcc-2.17.194.0+d312836f/MODULE_5f6d188e59db94aef5f0+165e9558/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.17.194.0+d312836f/0_REGISTRY/0.3.0.dev4/phi3/microsoft/Phi-3-mini-4k-instruct/32cf3abd57ca0217118c.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": "2.17.194.0+d312836f",
46
+ "num_cores_per_group": 1,
47
+ "on_device_sampling": false,
48
+ "optimum_neuron_version": "0.3.0.dev4",
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-2.17.194.0+d312836f/MODULE_04ccf0a18ef2a2487c5a+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_04ccf0a18ef2a2487c5a+bfe5714b/model.done ADDED
File without changes
neuronxcc-2.17.194.0+d312836f/MODULE_04ccf0a18ef2a2487c5a+bfe5714b/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:32a93db8e809a7d302bc166c199c7f558284713a8c9edc8a6fb1a33f3c355cc2
3
+ size 839488
neuronxcc-2.17.194.0+d312836f/MODULE_04ccf0a18ef2a2487c5a+bfe5714b/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:615bbda1bff6734a3f7528d608543250583b0d39a6885a331542783125b8950c
3
+ size 6216704
neuronxcc-2.17.194.0+d312836f/MODULE_0597e13911f7cd4171ab+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_0597e13911f7cd4171ab+431f5505/model.done ADDED
File without changes
neuronxcc-2.17.194.0+d312836f/MODULE_0597e13911f7cd4171ab+431f5505/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e89b9591280084034adda1cb99fdaac2f6b2360af3702e59e41e15065cc5419c
3
+ size 135999
neuronxcc-2.17.194.0+d312836f/MODULE_0597e13911f7cd4171ab+431f5505/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0784f60b6007c5ee21cecb5478f5496d0506e2972be221b59ed03a0f4c4ee35f
3
+ size 3974144
neuronxcc-2.17.194.0+d312836f/MODULE_5f6d188e59db94aef5f0+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_5f6d188e59db94aef5f0+165e9558/model.done ADDED
File without changes
neuronxcc-2.17.194.0+d312836f/MODULE_5f6d188e59db94aef5f0+165e9558/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ef7ab6d6e2622f6a6463b59d37f2e582429d8f2366f9ae536bea9331cf446a78
3
+ size 652730
neuronxcc-2.17.194.0+d312836f/MODULE_5f6d188e59db94aef5f0+165e9558/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a7681501b88b68766e1b663be3e7177858f1b69995cd2d82c5a657c21d46ec49
3
+ size 8582144
neuronxcc-2.17.194.0+d312836f/MODULE_5f6d188e59db94aef5f0+165e9558/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:55f46e69f2068b80f82b6005aaacc24295a08d465d88748eb806590d335e9eae
3
+ size 8719395