Synchronizing local compiler cache.
Browse files- .gitattributes +4 -0
- neuronxcc-2.17.194.0+d312836f/0_REGISTRY/0.3.0.dev1/phi3/microsoft/Phi-3-mini-4k-instruct/5b8ae963e4a07386d1f3.json +77 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_71d6f1724eb9ca6bbf34+431f5505/compile_flags.json +1 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_71d6f1724eb9ca6bbf34+431f5505/model.done +0 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_71d6f1724eb9ca6bbf34+431f5505/model.hlo_module.pb +3 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_71d6f1724eb9ca6bbf34+431f5505/model.neff +3 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_8d4a4e991c9ba3c051b5+165e9558/compile_flags.json +1 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_8d4a4e991c9ba3c051b5+165e9558/model.done +0 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_8d4a4e991c9ba3c051b5+165e9558/model.hlo_module.pb +3 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_8d4a4e991c9ba3c051b5+165e9558/model.neff +3 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_8d4a4e991c9ba3c051b5+165e9558/wrapped_neff.hlo +3 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_ddd8df8ccbf7e2239d19+bfe5714b/compile_flags.json +1 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_ddd8df8ccbf7e2239d19+bfe5714b/model.done +0 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_ddd8df8ccbf7e2239d19+bfe5714b/model.hlo_module.pb +3 -0
- neuronxcc-2.17.194.0+d312836f/MODULE_ddd8df8ccbf7e2239d19+bfe5714b/model.neff +3 -0
.gitattributes
CHANGED
@@ -8520,3 +8520,7 @@ neuronxcc-2.17.194.0+d312836f/MODULE_01abc9376daa952e0718+165e9558/model.neff fi
|
|
8520 |
neuronxcc-2.17.194.0+d312836f/MODULE_01abc9376daa952e0718+165e9558/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
8521 |
neuronxcc-2.17.194.0+d312836f/MODULE_51de7bf727a3a094effa+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
|
8522 |
neuronxcc-2.17.194.0+d312836f/MODULE_c808adc36dc289df1d8d+bfe5714b/model.neff filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
8520 |
neuronxcc-2.17.194.0+d312836f/MODULE_01abc9376daa952e0718+165e9558/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
8521 |
neuronxcc-2.17.194.0+d312836f/MODULE_51de7bf727a3a094effa+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
|
8522 |
neuronxcc-2.17.194.0+d312836f/MODULE_c808adc36dc289df1d8d+bfe5714b/model.neff filter=lfs diff=lfs merge=lfs -text
|
8523 |
+
neuronxcc-2.17.194.0+d312836f/MODULE_71d6f1724eb9ca6bbf34+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
|
8524 |
+
neuronxcc-2.17.194.0+d312836f/MODULE_8d4a4e991c9ba3c051b5+165e9558/model.neff filter=lfs diff=lfs merge=lfs -text
|
8525 |
+
neuronxcc-2.17.194.0+d312836f/MODULE_8d4a4e991c9ba3c051b5+165e9558/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
8526 |
+
neuronxcc-2.17.194.0+d312836f/MODULE_ddd8df8ccbf7e2239d19+bfe5714b/model.neff filter=lfs diff=lfs merge=lfs -text
|
neuronxcc-2.17.194.0+d312836f/0_REGISTRY/0.3.0.dev1/phi3/microsoft/Phi-3-mini-4k-instruct/5b8ae963e4a07386d1f3.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.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-2.17.194.0+d312836f/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-2.17.194.0+d312836f/MODULE_71d6f1724eb9ca6bbf34+431f5505/model.done
ADDED
File without changes
|
neuronxcc-2.17.194.0+d312836f/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-2.17.194.0+d312836f/MODULE_71d6f1724eb9ca6bbf34+431f5505/model.neff
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c3791490780e64f6f24c2be525649341ca9472f93b09a314ce6441f66e96f372
|
3 |
+
size 3974144
|
neuronxcc-2.17.194.0+d312836f/MODULE_8d4a4e991c9ba3c051b5+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_8d4a4e991c9ba3c051b5+165e9558/model.done
ADDED
File without changes
|
neuronxcc-2.17.194.0+d312836f/MODULE_8d4a4e991c9ba3c051b5+165e9558/model.hlo_module.pb
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:700b741ab6d230975421a5f5586b1eff563e80c3cfe196f38128861d87b2daa3
|
3 |
+
size 647962
|
neuronxcc-2.17.194.0+d312836f/MODULE_8d4a4e991c9ba3c051b5+165e9558/model.neff
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6a9ebcb7f06131fe03ca82fe43904b87d9d959c9692ccb29e8b7b84dc619e99a
|
3 |
+
size 10650624
|
neuronxcc-2.17.194.0+d312836f/MODULE_8d4a4e991c9ba3c051b5+165e9558/wrapped_neff.hlo
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d8b2faed5c11816517fb12018461300fbb1b65f248137575e8476155429f5009
|
3 |
+
size 10788002
|
neuronxcc-2.17.194.0+d312836f/MODULE_ddd8df8ccbf7e2239d19+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_ddd8df8ccbf7e2239d19+bfe5714b/model.done
ADDED
File without changes
|
neuronxcc-2.17.194.0+d312836f/MODULE_ddd8df8ccbf7e2239d19+bfe5714b/model.hlo_module.pb
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f1471300aaae4d1147af13017253b0d3b4b463f368087b93481273553b295c1e
|
3 |
+
size 835006
|
neuronxcc-2.17.194.0+d312836f/MODULE_ddd8df8ccbf7e2239d19+bfe5714b/model.neff
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:0626ccd085eceb184eda7f0019631866b9ba673e4bfb7e63133a7895fd367d12
|
3 |
+
size 6216704
|