Synchronizing local compiler cache.
Browse files- .gitattributes +3 -0
- neuronxcc-2.26.6360.0+6f180f47/0_REGISTRY/0.4.7.dev0/c54cbc73d9a74e547bf7ca1feb2b290b641ed261e32f7c08baba5633884f1298/4368e22aac9ef1cb4dc6.json +63 -0
- neuronxcc-2.26.6360.0+6f180f47/MODULE_319dcc32f1add4027ea9+12dd9119/compile_flags.json +1 -0
- neuronxcc-2.26.6360.0+6f180f47/MODULE_319dcc32f1add4027ea9+12dd9119/model.done +0 -0
- neuronxcc-2.26.6360.0+6f180f47/MODULE_319dcc32f1add4027ea9+12dd9119/model.hlo_module.pb +3 -0
- neuronxcc-2.26.6360.0+6f180f47/MODULE_319dcc32f1add4027ea9+12dd9119/model.neff +3 -0
- neuronxcc-2.26.6360.0+6f180f47/MODULE_c515cdf661866e372e4d+84337dd9/compile_flags.json +1 -0
- neuronxcc-2.26.6360.0+6f180f47/MODULE_c515cdf661866e372e4d+84337dd9/model.done +0 -0
- neuronxcc-2.26.6360.0+6f180f47/MODULE_c515cdf661866e372e4d+84337dd9/model.hlo_module.pb +3 -0
- neuronxcc-2.26.6360.0+6f180f47/MODULE_c515cdf661866e372e4d+84337dd9/model.neff +3 -0
- neuronxcc-2.26.6360.0+6f180f47/MODULE_c515cdf661866e372e4d+84337dd9/wrapped_neff.hlo +3 -0
.gitattributes
CHANGED
|
@@ -17819,3 +17819,6 @@ neuronxcc-2.26.6360.0+6f180f47/MODULE_b0694c190f21f974e8ef+84337dd9/wrapped_neff
|
|
| 17819 |
neuronxcc-2.26.6360.0+6f180f47/MODULE_009d38b6295dfb0b48a5+12dd9119/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 17820 |
neuronxcc-2.26.6360.0+6f180f47/MODULE_4b305156a7cf939164e5+84337dd9/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 17821 |
neuronxcc-2.26.6360.0+6f180f47/MODULE_4b305156a7cf939164e5+84337dd9/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
|
|
| 17819 |
neuronxcc-2.26.6360.0+6f180f47/MODULE_009d38b6295dfb0b48a5+12dd9119/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 17820 |
neuronxcc-2.26.6360.0+6f180f47/MODULE_4b305156a7cf939164e5+84337dd9/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 17821 |
neuronxcc-2.26.6360.0+6f180f47/MODULE_4b305156a7cf939164e5+84337dd9/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
| 17822 |
+
neuronxcc-2.26.6360.0+6f180f47/MODULE_319dcc32f1add4027ea9+12dd9119/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 17823 |
+
neuronxcc-2.26.6360.0+6f180f47/MODULE_c515cdf661866e372e4d+84337dd9/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 17824 |
+
neuronxcc-2.26.6360.0+6f180f47/MODULE_c515cdf661866e372e4d+84337dd9/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
neuronxcc-2.26.6360.0+6f180f47/0_REGISTRY/0.4.7.dev0/c54cbc73d9a74e547bf7ca1feb2b290b641ed261e32f7c08baba5633884f1298/4368e22aac9ef1cb4dc6.json
ADDED
|
@@ -0,0 +1,63 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_entry_class": "SingleModelCacheEntry",
|
| 3 |
+
"_model_id": "meta-llama/Meta-Llama-3.1-8B",
|
| 4 |
+
"_task": "text-generation",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"LlamaForCausalLM"
|
| 7 |
+
],
|
| 8 |
+
"attention_bias": false,
|
| 9 |
+
"attention_dropout": 0.0,
|
| 10 |
+
"dtype": "bfloat16",
|
| 11 |
+
"head_dim": 128,
|
| 12 |
+
"hidden_act": "silu",
|
| 13 |
+
"hidden_size": 4096,
|
| 14 |
+
"initializer_range": 0.02,
|
| 15 |
+
"intermediate_size": 14336,
|
| 16 |
+
"max_position_embeddings": 131072,
|
| 17 |
+
"mlp_bias": false,
|
| 18 |
+
"model_type": "llama",
|
| 19 |
+
"neuron": {
|
| 20 |
+
"_serialized_key": "NxDNeuronConfig",
|
| 21 |
+
"batch_size": 4,
|
| 22 |
+
"capacity_factor": null,
|
| 23 |
+
"checkpoint_id": "meta-llama/Meta-Llama-3.1-8B",
|
| 24 |
+
"checkpoint_revision": "d04e592bb4f6aa9cfee91e2e20afa771667e1d4b",
|
| 25 |
+
"continuous_batching": true,
|
| 26 |
+
"ep_degree": 1,
|
| 27 |
+
"fused_qkv": true,
|
| 28 |
+
"glu_mlp": true,
|
| 29 |
+
"local_ranks_size": 2,
|
| 30 |
+
"max_batch_size": 4,
|
| 31 |
+
"max_context_length": 4096,
|
| 32 |
+
"max_topk": 256,
|
| 33 |
+
"n_active_tokens": 4096,
|
| 34 |
+
"neuronxcc_version": "2.26.6360.0+6f180f47",
|
| 35 |
+
"on_device_sampling": true,
|
| 36 |
+
"optimum_neuron_version": "0.4.7.dev0",
|
| 37 |
+
"output_logits": false,
|
| 38 |
+
"pp_degree": 1,
|
| 39 |
+
"prefill_chunk_size": 1024,
|
| 40 |
+
"sequence_length": 4096,
|
| 41 |
+
"speculation_length": 0,
|
| 42 |
+
"start_rank_id": 0,
|
| 43 |
+
"target": "trn1",
|
| 44 |
+
"torch_dtype": "bfloat16",
|
| 45 |
+
"tp_degree": 2
|
| 46 |
+
},
|
| 47 |
+
"num_attention_heads": 32,
|
| 48 |
+
"num_hidden_layers": 32,
|
| 49 |
+
"num_key_value_heads": 8,
|
| 50 |
+
"pretraining_tp": 1,
|
| 51 |
+
"rms_norm_eps": 1e-05,
|
| 52 |
+
"rope_scaling": {
|
| 53 |
+
"factor": 8.0,
|
| 54 |
+
"high_freq_factor": 4.0,
|
| 55 |
+
"low_freq_factor": 1.0,
|
| 56 |
+
"original_max_position_embeddings": 8192,
|
| 57 |
+
"rope_type": "llama3"
|
| 58 |
+
},
|
| 59 |
+
"rope_theta": 500000.0,
|
| 60 |
+
"tie_word_embeddings": false,
|
| 61 |
+
"use_cache": true,
|
| 62 |
+
"vocab_size": 128256
|
| 63 |
+
}
|
neuronxcc-2.26.6360.0+6f180f47/MODULE_319dcc32f1add4027ea9+12dd9119/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", "--verbose=35", "--logfile=/tmp/nxd_model/chunked_prefill/_tp0_bk0/log-neuron-cc.txt"]
|
neuronxcc-2.26.6360.0+6f180f47/MODULE_319dcc32f1add4027ea9+12dd9119/model.done
ADDED
|
File without changes
|
neuronxcc-2.26.6360.0+6f180f47/MODULE_319dcc32f1add4027ea9+12dd9119/model.hlo_module.pb
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c440efd75b0c2168bf5941c8de2f43343533d3bdf5c1127d48a1481a9cf5034f
|
| 3 |
+
size 4569650
|
neuronxcc-2.26.6360.0+6f180f47/MODULE_319dcc32f1add4027ea9+12dd9119/model.neff
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dee259e269ba271909df72d058b99629a3a677ae9be2b8a347b84a661f3e3e04
|
| 3 |
+
size 4680704
|
neuronxcc-2.26.6360.0+6f180f47/MODULE_c515cdf661866e372e4d+84337dd9/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", "--verbose=35", "--logfile=/tmp/nxd_model/token_generation/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
|
neuronxcc-2.26.6360.0+6f180f47/MODULE_c515cdf661866e372e4d+84337dd9/model.done
ADDED
|
File without changes
|
neuronxcc-2.26.6360.0+6f180f47/MODULE_c515cdf661866e372e4d+84337dd9/model.hlo_module.pb
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:12fd6499b9e73348dc70290183d686646a677419b598064f87bcbaba8896c81e
|
| 3 |
+
size 3318639
|
neuronxcc-2.26.6360.0+6f180f47/MODULE_c515cdf661866e372e4d+84337dd9/model.neff
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ec0f9f7eb6d1b59b12612526db8b752eafe67feb7ca8fa2f27ada0fb9302e65d
|
| 3 |
+
size 10937344
|
neuronxcc-2.26.6360.0+6f180f47/MODULE_c515cdf661866e372e4d+84337dd9/wrapped_neff.hlo
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0647c57e937ccf6bd6c1876dd1469ad326460809ac3e969b3a1b83e35a5533ff
|
| 3 |
+
size 11085274
|