Synchronizing local compiler cache.
Browse files- .gitattributes +4 -0
- neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.0.dev4/phi3/microsoft/Phi-3-mini-4k-instruct/6d4d2a0dd1dabb55e29d.json +77 -0
- neuronxcc-2.19.8089.0+8ab9f450/MODULE_0597e13911f7cd4171ab+431f5505/compile_flags.json +1 -0
- neuronxcc-2.19.8089.0+8ab9f450/MODULE_0597e13911f7cd4171ab+431f5505/model.done +0 -0
- neuronxcc-2.19.8089.0+8ab9f450/MODULE_0597e13911f7cd4171ab+431f5505/model.hlo_module.pb +3 -0
- neuronxcc-2.19.8089.0+8ab9f450/MODULE_0597e13911f7cd4171ab+431f5505/model.neff +3 -0
- neuronxcc-2.19.8089.0+8ab9f450/MODULE_368872307eefe85b30ca+ed72d204/compile_flags.json +1 -0
- neuronxcc-2.19.8089.0+8ab9f450/MODULE_368872307eefe85b30ca+ed72d204/model.done +0 -0
- neuronxcc-2.19.8089.0+8ab9f450/MODULE_368872307eefe85b30ca+ed72d204/model.hlo_module.pb +3 -0
- neuronxcc-2.19.8089.0+8ab9f450/MODULE_368872307eefe85b30ca+ed72d204/model.neff +3 -0
- neuronxcc-2.19.8089.0+8ab9f450/MODULE_85679c6b0c96942cd024+a9d440f5/compile_flags.json +1 -0
- neuronxcc-2.19.8089.0+8ab9f450/MODULE_85679c6b0c96942cd024+a9d440f5/model.done +0 -0
- neuronxcc-2.19.8089.0+8ab9f450/MODULE_85679c6b0c96942cd024+a9d440f5/model.hlo_module.pb +3 -0
- neuronxcc-2.19.8089.0+8ab9f450/MODULE_85679c6b0c96942cd024+a9d440f5/model.neff +3 -0
- neuronxcc-2.19.8089.0+8ab9f450/MODULE_85679c6b0c96942cd024+a9d440f5/wrapped_neff.hlo +3 -0
.gitattributes
CHANGED
@@ -3624,3 +3624,7 @@ neuronxcc-2.19.8089.0+8ab9f450/MODULE_fe5cc26344fc85d3582c+431f5505/model.neff f
|
|
3624 |
neuronxcc-2.19.8089.0+8ab9f450/MODULE_313796881a2973827e21+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
|
3625 |
neuronxcc-2.19.8089.0+8ab9f450/MODULE_484d20af08b29ebeab90+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
|
3626 |
neuronxcc-2.19.8089.0+8ab9f450/MODULE_484d20af08b29ebeab90+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
3624 |
neuronxcc-2.19.8089.0+8ab9f450/MODULE_313796881a2973827e21+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
|
3625 |
neuronxcc-2.19.8089.0+8ab9f450/MODULE_484d20af08b29ebeab90+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
|
3626 |
neuronxcc-2.19.8089.0+8ab9f450/MODULE_484d20af08b29ebeab90+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
3627 |
+
neuronxcc-2.19.8089.0+8ab9f450/MODULE_0597e13911f7cd4171ab+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
|
3628 |
+
neuronxcc-2.19.8089.0+8ab9f450/MODULE_368872307eefe85b30ca+ed72d204/model.neff filter=lfs diff=lfs merge=lfs -text
|
3629 |
+
neuronxcc-2.19.8089.0+8ab9f450/MODULE_85679c6b0c96942cd024+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
|
3630 |
+
neuronxcc-2.19.8089.0+8ab9f450/MODULE_85679c6b0c96942cd024+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.0.dev4/phi3/microsoft/Phi-3-mini-4k-instruct/6d4d2a0dd1dabb55e29d.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": "2.19.8089.0+8ab9f450",
|
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-2.19.8089.0+8ab9f450/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-2.19.8089.0+8ab9f450/MODULE_0597e13911f7cd4171ab+431f5505/model.done
ADDED
File without changes
|
neuronxcc-2.19.8089.0+8ab9f450/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-2.19.8089.0+8ab9f450/MODULE_0597e13911f7cd4171ab+431f5505/model.neff
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:be4d5634f01e311389b513d77a8cac73a7a5b79532042bfb18678b2ee7eeb467
|
3 |
+
size 3974144
|
neuronxcc-2.19.8089.0+8ab9f450/MODULE_368872307eefe85b30ca+ed72d204/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-strided-dma ", "-O2", "--lnc=1", "--logfile=/tmp/nxd_model/context_encoding_model/_tp0_bk0/log-neuron-cc.txt"]
|
neuronxcc-2.19.8089.0+8ab9f450/MODULE_368872307eefe85b30ca+ed72d204/model.done
ADDED
File without changes
|
neuronxcc-2.19.8089.0+8ab9f450/MODULE_368872307eefe85b30ca+ed72d204/model.hlo_module.pb
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:431e29d2bb064ef861c01ace6150c1b088306847023b1c9b69d5d6a167fe318f
|
3 |
+
size 862528
|
neuronxcc-2.19.8089.0+8ab9f450/MODULE_368872307eefe85b30ca+ed72d204/model.neff
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9265e58445fa03e30ca19059d36c76ddbd960cf18b9796d36d6842390fea0795
|
3 |
+
size 5786624
|
neuronxcc-2.19.8089.0+8ab9f450/MODULE_85679c6b0c96942cd024+a9d440f5/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-strided-dma ", "-O2", "--lnc=1", "--logfile=/tmp/nxd_model/token_generation_model/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
|
neuronxcc-2.19.8089.0+8ab9f450/MODULE_85679c6b0c96942cd024+a9d440f5/model.done
ADDED
File without changes
|
neuronxcc-2.19.8089.0+8ab9f450/MODULE_85679c6b0c96942cd024+a9d440f5/model.hlo_module.pb
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:fa569dbe555a0209c2345d4ed6a1fc32945b47856780d8d599e8a6a1d40059d3
|
3 |
+
size 652730
|
neuronxcc-2.19.8089.0+8ab9f450/MODULE_85679c6b0c96942cd024+a9d440f5/model.neff
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:cb237fb8bb88288a82440f8d894563d6326de6269841d280119c6a1deba7f9b9
|
3 |
+
size 8817664
|
neuronxcc-2.19.8089.0+8ab9f450/MODULE_85679c6b0c96942cd024+a9d440f5/wrapped_neff.hlo
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f57b8797a60325ecec4b92c5e50d97fe729a4cd26b39f5a98edbbf16e5ab76cf
|
3 |
+
size 8954915
|