tengomucho HF Staff commited on
Commit
b409e5c
·
verified ·
1 Parent(s): aefca07

Synchronizing local compiler cache.

Browse files
.gitattributes CHANGED
@@ -3932,3 +3932,6 @@ neuronxcc-2.19.8089.0+8ab9f450/MODULE_7468246fa1f76cb36408+ca355898/wrapped_neff
3932
  neuronxcc-2.19.8089.0+8ab9f450/MODULE_ac21214b87b47765781a+c2248236/model.neff filter=lfs diff=lfs merge=lfs -text
3933
  neuronxcc-2.19.8089.0+8ab9f450/MODULE_b0e34110163a7a5f45af+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
3934
  neuronxcc-2.19.8089.0+8ab9f450/MODULE_b0e34110163a7a5f45af+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
 
 
 
 
3932
  neuronxcc-2.19.8089.0+8ab9f450/MODULE_ac21214b87b47765781a+c2248236/model.neff filter=lfs diff=lfs merge=lfs -text
3933
  neuronxcc-2.19.8089.0+8ab9f450/MODULE_b0e34110163a7a5f45af+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
3934
  neuronxcc-2.19.8089.0+8ab9f450/MODULE_b0e34110163a7a5f45af+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
3935
+ neuronxcc-2.19.8089.0+8ab9f450/MODULE_2a5016954fa7d0cd1b4a+c2248236/model.neff filter=lfs diff=lfs merge=lfs -text
3936
+ neuronxcc-2.19.8089.0+8ab9f450/MODULE_3224f0ff6a6f0dea53dc+ca355898/model.neff filter=lfs diff=lfs merge=lfs -text
3937
+ neuronxcc-2.19.8089.0+8ab9f450/MODULE_3224f0ff6a6f0dea53dc+ca355898/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.19.8089.0+8ab9f450/0_REGISTRY/0.3.1.dev0/gpt_oss/plop-internal/gpt-oss-20b-trfs-0804/0410e02658acf144c0b4.json ADDED
@@ -0,0 +1,120 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "plop-internal/gpt-oss-20b-trfs-0804",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "GptOssForCausalLM"
7
+ ],
8
+ "attention_bias": true,
9
+ "attention_dropout": 0.0,
10
+ "experts_per_token": 4,
11
+ "head_dim": 64,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 2880,
14
+ "initial_context_length": 4096,
15
+ "initializer_range": 0.02,
16
+ "intermediate_size": 2880,
17
+ "layer_types": [
18
+ "sliding_attention",
19
+ "full_attention",
20
+ "sliding_attention",
21
+ "full_attention",
22
+ "sliding_attention",
23
+ "full_attention",
24
+ "sliding_attention",
25
+ "full_attention",
26
+ "sliding_attention",
27
+ "full_attention",
28
+ "sliding_attention",
29
+ "full_attention",
30
+ "sliding_attention",
31
+ "full_attention",
32
+ "sliding_attention",
33
+ "full_attention",
34
+ "sliding_attention",
35
+ "full_attention",
36
+ "sliding_attention",
37
+ "full_attention",
38
+ "sliding_attention",
39
+ "full_attention",
40
+ "sliding_attention",
41
+ "full_attention"
42
+ ],
43
+ "max_position_embeddings": 131072,
44
+ "mlp_bias": false,
45
+ "model_type": "gpt_oss",
46
+ "neuron": {
47
+ "_serialized_key": "NxDNeuronConfig",
48
+ "async_mode": false,
49
+ "attn_kernel_enabled": false,
50
+ "batch_size": 1,
51
+ "capacity_factor": null,
52
+ "cc_pipeline_tiling_factor": 2,
53
+ "checkpoint_id": "plop-internal/gpt-oss-20b-trfs-0804",
54
+ "checkpoint_revision": "8964acebadbe440e1f70a40383a95d7f9eec20bc",
55
+ "continuous_batching": false,
56
+ "enable_bucketing": false,
57
+ "ep_degree": 1,
58
+ "flash_decoding_enabled": false,
59
+ "fused_qkv": false,
60
+ "glu_mlp": true,
61
+ "is_chunked_prefill": false,
62
+ "local_ranks_size": 8,
63
+ "logical_nc_config": 1,
64
+ "max_batch_size": 1,
65
+ "max_context_length": 256,
66
+ "max_topk": 256,
67
+ "mlp_kernel_enabled": false,
68
+ "mlp_kernel_fuse_residual_add": false,
69
+ "n_active_tokens": 256,
70
+ "neuronxcc_version": "2.19.8089.0+8ab9f450",
71
+ "num_cores_per_group": 1,
72
+ "on_device_sampling": false,
73
+ "optimum_neuron_version": "0.3.1.dev0",
74
+ "output_logits": false,
75
+ "padding_side": "right",
76
+ "pp_degree": 1,
77
+ "qk_layernorm": false,
78
+ "qkv_kernel_enabled": false,
79
+ "rpl_reduce_dtype": "bfloat16",
80
+ "sequence_length": 256,
81
+ "sequence_parallel_enabled": false,
82
+ "speculation_length": 0,
83
+ "start_rank_id": 0,
84
+ "target": null,
85
+ "torch_dtype": "bfloat16",
86
+ "tp_degree": 8,
87
+ "vocab_parallel": false
88
+ },
89
+ "num_attention_heads": 64,
90
+ "num_experts_per_tok": 4,
91
+ "num_hidden_layers": 24,
92
+ "num_key_value_heads": 8,
93
+ "num_local_experts": 32,
94
+ "output_router_logits": false,
95
+ "quantization_config": {
96
+ "modules_to_not_convert": [
97
+ "model.layers.*.self_attn",
98
+ "model.layers.*.mlp.router",
99
+ "model.embed_tokens",
100
+ "lm_head"
101
+ ],
102
+ "quant_method": "mxfp4"
103
+ },
104
+ "rms_norm_eps": 1e-05,
105
+ "rope_scaling": {
106
+ "beta_fast": 32.0,
107
+ "beta_slow": 1.0,
108
+ "factor": 32.0,
109
+ "original_max_position_embeddings": 4096,
110
+ "rope_type": "yarn",
111
+ "truncate": false
112
+ },
113
+ "rope_theta": 150000,
114
+ "router_aux_loss_coef": 0.9,
115
+ "sliding_window": 128,
116
+ "swiglu_limit": 7.0,
117
+ "tie_word_embeddings": false,
118
+ "use_cache": true,
119
+ "vocab_size": 201088
120
+ }
neuronxcc-2.19.8089.0+8ab9f450/MODULE_2a5016954fa7d0cd1b4a+c2248236/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--enable-saturate-infinity", "--enable-mixed-precision-accumulation", "--model-type", "transformer", "-O1", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2", "--auto-cast=none", "--internal-enable-dge-levels", "vector_dynamic_offsets", "--logfile=/tmp/nxd_model/context_encoding_model/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.19.8089.0+8ab9f450/MODULE_2a5016954fa7d0cd1b4a+c2248236/model.done ADDED
File without changes
neuronxcc-2.19.8089.0+8ab9f450/MODULE_2a5016954fa7d0cd1b4a+c2248236/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:989e1c4745a465272ac6c284ed143a111a086f6b8f0402d1cf78922ec045ccdc
3
+ size 12833532
neuronxcc-2.19.8089.0+8ab9f450/MODULE_2a5016954fa7d0cd1b4a+c2248236/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a76d8af43eb4e14a0093975a8980f7a927c0be8a736d0d95750c4486eab5f5e8
3
+ size 6841344
neuronxcc-2.19.8089.0+8ab9f450/MODULE_3224f0ff6a6f0dea53dc+ca355898/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--enable-saturate-infinity", "--enable-mixed-precision-accumulation", "--model-type", "transformer", "-O1", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2", "--auto-cast=none", "--internal-enable-dge-levels", "vector_dynamic_offsets", "--logfile=/tmp/nxd_model/token_generation_model/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
neuronxcc-2.19.8089.0+8ab9f450/MODULE_3224f0ff6a6f0dea53dc+ca355898/model.done ADDED
File without changes
neuronxcc-2.19.8089.0+8ab9f450/MODULE_3224f0ff6a6f0dea53dc+ca355898/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:09554eaec6559839de07d2615698cfb41918c284c3e0accd07c336c732dd73c2
3
+ size 1011147
neuronxcc-2.19.8089.0+8ab9f450/MODULE_3224f0ff6a6f0dea53dc+ca355898/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f37185c9d12ce9f804459222c0aa7edf5c9f95b836e51718a7bab31c57c4978e
3
+ size 2192384
neuronxcc-2.19.8089.0+8ab9f450/MODULE_3224f0ff6a6f0dea53dc+ca355898/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9dc7db5538518f887725a19b256205b57b173561fb94c258f45578a30523d5e9
3
+ size 2350817