Hugging Face's logo

aws-neuron
/
optimum-neuron-cache

Model card Files
xet
Community
598
dacorvo HF Staff commited on
Commit
89de499
·
verified ·
1 Parent(s): b34c487

Synchronizing local compiler cache.

Browse files
Files changed (26) hide show
  1. .gitattributes +6 -0
  2. neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/qwen3/Qwen/Qwen3-8B/d7608dc7ce9ca58d8b1b.json +95 -0
  3. neuronxcc-2.21.18209.0+043b1bf7/MODULE_084b0ca25510be3eae0e+ed72d204/compile_flags.json +1 -0
  4. neuronxcc-2.21.18209.0+043b1bf7/MODULE_084b0ca25510be3eae0e+ed72d204/model.hlo_module.pb +3 -0
  5. neuronxcc-2.21.18209.0+043b1bf7/MODULE_084b0ca25510be3eae0e+ed72d204/model.log +3 -0
  6. neuronxcc-2.21.18209.0+043b1bf7/MODULE_63b08aa574a103e133be+a9d440f5/compile_flags.json +1 -0
  7. neuronxcc-2.21.18209.0+043b1bf7/MODULE_63b08aa574a103e133be+a9d440f5/model.done +0 -0
  8. neuronxcc-2.21.18209.0+043b1bf7/MODULE_63b08aa574a103e133be+a9d440f5/model.hlo_module.pb +3 -0
  9. neuronxcc-2.21.18209.0+043b1bf7/MODULE_63b08aa574a103e133be+a9d440f5/model.neff +3 -0
  10. neuronxcc-2.21.18209.0+043b1bf7/MODULE_63b08aa574a103e133be+a9d440f5/wrapped_neff.hlo +3 -0
  11. neuronxcc-2.21.18209.0+043b1bf7/MODULE_7585b7b81ecc283af772+a9d440f5/compile_flags.json +1 -0
  12. neuronxcc-2.21.18209.0+043b1bf7/MODULE_7585b7b81ecc283af772+a9d440f5/model.done +0 -0
  13. neuronxcc-2.21.18209.0+043b1bf7/MODULE_7585b7b81ecc283af772+a9d440f5/model.hlo_module.pb +3 -0
  14. neuronxcc-2.21.18209.0+043b1bf7/MODULE_7585b7b81ecc283af772+a9d440f5/model.neff +3 -0
  15. neuronxcc-2.21.18209.0+043b1bf7/MODULE_7585b7b81ecc283af772+a9d440f5/wrapped_neff.hlo +3 -0
  16. neuronxcc-2.21.18209.0+043b1bf7/MODULE_bae931052fc7117dae12+a9d440f5/compile_flags.json +1 -0
  17. neuronxcc-2.21.18209.0+043b1bf7/MODULE_bae931052fc7117dae12+a9d440f5/model.done +0 -0
  18. neuronxcc-2.21.18209.0+043b1bf7/MODULE_bae931052fc7117dae12+a9d440f5/model.hlo_module.pb +3 -0
  19. neuronxcc-2.21.18209.0+043b1bf7/MODULE_bae931052fc7117dae12+a9d440f5/model.neff +3 -0
  20. neuronxcc-2.21.18209.0+043b1bf7/MODULE_bae931052fc7117dae12+a9d440f5/wrapped_neff.hlo +3 -0
  21. neuronxcc-2.21.18209.0+043b1bf7/MODULE_fdf838eca836e25ae600+ed72d204/compile_flags.json +1 -0
  22. neuronxcc-2.21.18209.0+043b1bf7/MODULE_fdf838eca836e25ae600+ed72d204/model.hlo_module.pb +3 -0
  23. neuronxcc-2.21.18209.0+043b1bf7/MODULE_fdf838eca836e25ae600+ed72d204/model.log +3 -0
  24. neuronxcc-2.21.18209.0+043b1bf7/MODULE_ff564f7b9f2edb395631+ed72d204/compile_flags.json +1 -0
  25. neuronxcc-2.21.18209.0+043b1bf7/MODULE_ff564f7b9f2edb395631+ed72d204/model.hlo_module.pb +3 -0
  26. neuronxcc-2.21.18209.0+043b1bf7/MODULE_ff564f7b9f2edb395631+ed72d204/model.log +3 -0
.gitattributes CHANGED
@@ -11840,3 +11840,9 @@ neuronxcc-2.21.18209.0+043b1bf7/MODULE_9f1c978a730025d88f0e+877608f3/model.neff
11840
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_3a0b53d0c2e168ed7240+747527b0/model.neff filter=lfs diff=lfs merge=lfs -text
11841
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_3a0b53d0c2e168ed7240+747527b0/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
11842
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_793c288702976d9a9911+877608f3/model.neff filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
 
11840
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_3a0b53d0c2e168ed7240+747527b0/model.neff filter=lfs diff=lfs merge=lfs -text
11841
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_3a0b53d0c2e168ed7240+747527b0/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
11842
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_793c288702976d9a9911+877608f3/model.neff filter=lfs diff=lfs merge=lfs -text
11843
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_63b08aa574a103e133be+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
11844
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_63b08aa574a103e133be+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
11845
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_7585b7b81ecc283af772+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
11846
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_7585b7b81ecc283af772+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
11847
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_bae931052fc7117dae12+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
11848
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_bae931052fc7117dae12+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0/qwen3/Qwen/Qwen3-8B/d7608dc7ce9ca58d8b1b.json ADDED
@@ -0,0 +1,95 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "Qwen/Qwen3-8B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Qwen3ForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 4096,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 12288,
15
+ "layer_types": [
16
+ "full_attention",
17
+ "full_attention",
18
+ "full_attention",
19
+ "full_attention",
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention",
45
+ "full_attention",
46
+ "full_attention",
47
+ "full_attention",
48
+ "full_attention",
49
+ "full_attention",
50
+ "full_attention",
51
+ "full_attention"
52
+ ],
53
+ "max_position_embeddings": 40960,
54
+ "max_window_layers": 36,
55
+ "model_type": "qwen3",
56
+ "neuron": {
57
+ "_serialized_key": "NxDNeuronConfig",
58
+ "batch_size": 1,
59
+ "capacity_factor": null,
60
+ "checkpoint_id": "Qwen/Qwen3-8B",
61
+ "checkpoint_revision": "b968826d9c46dd6066d109eabc6255188de91218",
62
+ "continuous_batching": false,
63
+ "enable_bucketing": false,
64
+ "ep_degree": 1,
65
+ "fused_qkv": true,
66
+ "glu_mlp": true,
67
+ "local_ranks_size": 2,
68
+ "max_batch_size": 1,
69
+ "max_context_length": 4096,
70
+ "max_topk": 256,
71
+ "n_active_tokens": 4096,
72
+ "neuronxcc_version": "2.21.18209.0+043b1bf7",
73
+ "on_device_sampling": true,
74
+ "optimum_neuron_version": "0.4.0",
75
+ "output_logits": false,
76
+ "pp_degree": 1,
77
+ "sequence_length": 4096,
78
+ "speculation_length": 0,
79
+ "start_rank_id": 0,
80
+ "target": "trn1",
81
+ "torch_dtype": "bfloat16",
82
+ "tp_degree": 2
83
+ },
84
+ "num_attention_heads": 32,
85
+ "num_hidden_layers": 36,
86
+ "num_key_value_heads": 8,
87
+ "rms_norm_eps": 1e-06,
88
+ "rope_scaling": null,
89
+ "rope_theta": 1000000,
90
+ "sliding_window": null,
91
+ "tie_word_embeddings": false,
92
+ "use_cache": true,
93
+ "use_sliding_window": false,
94
+ "vocab_size": 151936
95
+ }
neuronxcc-2.21.18209.0+043b1bf7/MODULE_084b0ca25510be3eae0e+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-2.21.18209.0+043b1bf7/MODULE_084b0ca25510be3eae0e+ed72d204/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cbe936df2dd3e7c0552c88652dd566f94ae8eafb06b7049095032f9603622a0e
3
+ size 106099506
neuronxcc-2.21.18209.0+043b1bf7/MODULE_084b0ca25510be3eae0e+ed72d204/model.log ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ Failed compilation with ['neuronx-cc', 'compile', '--framework=XLA', '/tmp/nxd_model/context_encoding_model/_tp0_bk0/model.MODULE_084b0ca25510be3eae0e+ed72d204.hlo_module.pb', '--output', '/tmp/nxd_model/context_encoding_model/_tp0_bk0/model.MODULE_084b0ca25510be3eae0e+ed72d204.neff', '--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', '--verbose=35']: [XCG815] Estimated peak HBM usage (19.849GB) exceeds 16GB. Neff might be unable to load on chip. If you believe this estimation to be inaccurate, you can disable the check using: `--internal-backend-options=' --disable-hbm-usage-check '` - Please open a support ticket at https://github.com/aws-neuron/aws-neuron-sdk/issues/new. You may also be able to obtain more information using the 'XLA_IR_DEBUG' and 'XLA_HLO_DEBUG' environment variables.
2
+ 2025-10-13T11:49:46Z Non-signal exit. Backend exited with code 1 and stderr: [XCG815] Estimated peak HBM usage (19.849GB) exceeds 16GB. Neff might be unable to load on chip. If you believe this estimation to be inaccurate, you can disable the check using: `--internal-backend-options=' --disable-hbm-usage-check '` - Please open a support ticket at https://github.com/aws-neuron/aws-neuron-sdk/issues/new. You may also be able to obtain more information using the 'XLA_IR_DEBUG' and 'XLA_HLO_DEBUG' environment variables.
3
+
neuronxcc-2.21.18209.0+043b1bf7/MODULE_63b08aa574a103e133be+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-2.21.18209.0+043b1bf7/MODULE_63b08aa574a103e133be+a9d440f5/model.done ADDED
File without changes
neuronxcc-2.21.18209.0+043b1bf7/MODULE_63b08aa574a103e133be+a9d440f5/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:21db4a0343706c21e048588561690b2061c787f1e4456d5abdd455064df96102
3
+ size 106127270
neuronxcc-2.21.18209.0+043b1bf7/MODULE_63b08aa574a103e133be+a9d440f5/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8525b10b7b1e2ee02a362c164a364ea9ac01d1f09c3b8a1862168f36a95ff0c2
3
+ size 15197184
neuronxcc-2.21.18209.0+043b1bf7/MODULE_63b08aa574a103e133be+a9d440f5/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d5658d3aa5ed4dd10106028e20617951b8c4515086735196b94908b3059e5752
3
+ size 15506153
neuronxcc-2.21.18209.0+043b1bf7/MODULE_7585b7b81ecc283af772+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-2.21.18209.0+043b1bf7/MODULE_7585b7b81ecc283af772+a9d440f5/model.done ADDED
File without changes
neuronxcc-2.21.18209.0+043b1bf7/MODULE_7585b7b81ecc283af772+a9d440f5/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:32e98bb691e366aa22d2d9dd2fba85706bd818c47344c0db74a983fa34b92209
3
+ size 103993280
neuronxcc-2.21.18209.0+043b1bf7/MODULE_7585b7b81ecc283af772+a9d440f5/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7b99d837a49291f3f272568a63bffef06ad4b129790ae2cbf99c22ef528ad46f
3
+ size 9882624
neuronxcc-2.21.18209.0+043b1bf7/MODULE_7585b7b81ecc283af772+a9d440f5/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cb901e10625a3eb094190d4f0ba055841675a2167f630e11531a72db0a567d4a
3
+ size 10191593
neuronxcc-2.21.18209.0+043b1bf7/MODULE_bae931052fc7117dae12+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-2.21.18209.0+043b1bf7/MODULE_bae931052fc7117dae12+a9d440f5/model.done ADDED
File without changes
neuronxcc-2.21.18209.0+043b1bf7/MODULE_bae931052fc7117dae12+a9d440f5/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fb344c672a010536aa058ddd9a19cd7de40cb29e0b785ac541c8dd2fc630085f
3
+ size 102926602
neuronxcc-2.21.18209.0+043b1bf7/MODULE_bae931052fc7117dae12+a9d440f5/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6b6e713335f2db8daa35894784cef45b53887ac954d0d59885127453500bdca3
3
+ size 7128064
neuronxcc-2.21.18209.0+043b1bf7/MODULE_bae931052fc7117dae12+a9d440f5/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:905da3ec1ebfaeb665f2806414ad6bf492f62b129781f762265293a80f8951d1
3
+ size 7437033
neuronxcc-2.21.18209.0+043b1bf7/MODULE_fdf838eca836e25ae600+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-2.21.18209.0+043b1bf7/MODULE_fdf838eca836e25ae600+ed72d204/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:114e077a57d5c1bd85c3f7cdbbe8c5da34c6e1875c9db46e4109da6a55d7f464
3
+ size 106099506
neuronxcc-2.21.18209.0+043b1bf7/MODULE_fdf838eca836e25ae600+ed72d204/model.log ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ Failed compilation with ['neuronx-cc', 'compile', '--framework=XLA', '/tmp/nxd_model/context_encoding_model/_tp0_bk0/model.MODULE_fdf838eca836e25ae600+ed72d204.hlo_module.pb', '--output', '/tmp/nxd_model/context_encoding_model/_tp0_bk0/model.MODULE_fdf838eca836e25ae600+ed72d204.neff', '--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', '--verbose=35']: [XCG815] Estimated peak HBM usage (19.099GB) exceeds 16GB. Neff might be unable to load on chip. If you believe this estimation to be inaccurate, you can disable the check using: `--internal-backend-options=' --disable-hbm-usage-check '` - Please open a support ticket at https://github.com/aws-neuron/aws-neuron-sdk/issues/new. You may also be able to obtain more information using the 'XLA_IR_DEBUG' and 'XLA_HLO_DEBUG' environment variables.
2
+ 2025-10-13T11:32:53Z Non-signal exit. Backend exited with code 1 and stderr: [XCG815] Estimated peak HBM usage (19.099GB) exceeds 16GB. Neff might be unable to load on chip. If you believe this estimation to be inaccurate, you can disable the check using: `--internal-backend-options=' --disable-hbm-usage-check '` - Please open a support ticket at https://github.com/aws-neuron/aws-neuron-sdk/issues/new. You may also be able to obtain more information using the 'XLA_IR_DEBUG' and 'XLA_HLO_DEBUG' environment variables.
3
+
neuronxcc-2.21.18209.0+043b1bf7/MODULE_ff564f7b9f2edb395631+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-2.21.18209.0+043b1bf7/MODULE_ff564f7b9f2edb395631+ed72d204/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cb1740c916e0af0675009dd8dfb70af5f8acc5e17bfd57ed1c96a30c750c671a
3
+ size 106099506
neuronxcc-2.21.18209.0+043b1bf7/MODULE_ff564f7b9f2edb395631+ed72d204/model.log ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ Failed compilation with ['neuronx-cc', 'compile', '--framework=XLA', '/tmp/nxd_model/context_encoding_model/_tp0_bk0/model.MODULE_ff564f7b9f2edb395631+ed72d204.hlo_module.pb', '--output', '/tmp/nxd_model/context_encoding_model/_tp0_bk0/model.MODULE_ff564f7b9f2edb395631+ed72d204.neff', '--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', '--verbose=35']: [XCG815] Estimated peak HBM usage (21.348GB) exceeds 16GB. Neff might be unable to load on chip. If you believe this estimation to be inaccurate, you can disable the check using: `--internal-backend-options=' --disable-hbm-usage-check '` - Please open a support ticket at https://github.com/aws-neuron/aws-neuron-sdk/issues/new. You may also be able to obtain more information using the 'XLA_IR_DEBUG' and 'XLA_HLO_DEBUG' environment variables.
2
+ 2025-10-13T12:10:29Z Non-signal exit. Backend exited with code 1 and stderr: [XCG815] Estimated peak HBM usage (21.348GB) exceeds 16GB. Neff might be unable to load on chip. If you believe this estimation to be inaccurate, you can disable the check using: `--internal-backend-options=' --disable-hbm-usage-check '` - Please open a support ticket at https://github.com/aws-neuron/aws-neuron-sdk/issues/new. You may also be able to obtain more information using the 'XLA_IR_DEBUG' and 'XLA_HLO_DEBUG' environment variables.
3
+