Synchronizing local compiler cache.
Browse files- .gitattributes +3 -0
- neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/meta-llama/Llama-3.1-8B/3b69da34b19e6d21ec06.json +62 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_3fba20491a2dd90d7fd7+ac10809c/compile_flags.json +1 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_3fba20491a2dd90d7fd7+ac10809c/model.done +0 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_3fba20491a2dd90d7fd7+ac10809c/model.hlo_module.pb +3 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_3fba20491a2dd90d7fd7+ac10809c/model.neff +3 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_3fba20491a2dd90d7fd7+ac10809c/wrapped_neff.hlo +3 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_ebdaf819ea1c1978f2f3+a32116a7/compile_flags.json +1 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_ebdaf819ea1c1978f2f3+a32116a7/model.done +0 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_ebdaf819ea1c1978f2f3+a32116a7/model.hlo_module.pb +3 -0
- neuronxcc-2.21.33363.0+82129205/MODULE_ebdaf819ea1c1978f2f3+a32116a7/model.neff +3 -0
.gitattributes
CHANGED
|
@@ -15342,3 +15342,6 @@ neuronxcc-2.21.33363.0+82129205/MODULE_9c964442c358261ab331+24129607/model.neff
|
|
| 15342 |
neuronxcc-2.21.33363.0+82129205/MODULE_24ce1d8a8a1d46cf5874+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 15343 |
neuronxcc-2.21.33363.0+82129205/MODULE_b2a120f544d3e85592b1+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 15344 |
neuronxcc-2.21.33363.0+82129205/MODULE_b2a120f544d3e85592b1+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
|
|
| 15342 |
neuronxcc-2.21.33363.0+82129205/MODULE_24ce1d8a8a1d46cf5874+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 15343 |
neuronxcc-2.21.33363.0+82129205/MODULE_b2a120f544d3e85592b1+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 15344 |
neuronxcc-2.21.33363.0+82129205/MODULE_b2a120f544d3e85592b1+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
| 15345 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_3fba20491a2dd90d7fd7+ac10809c/model.neff filter=lfs diff=lfs merge=lfs -text
|
| 15346 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_3fba20491a2dd90d7fd7+ac10809c/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
|
| 15347 |
+
neuronxcc-2.21.33363.0+82129205/MODULE_ebdaf819ea1c1978f2f3+a32116a7/model.neff filter=lfs diff=lfs merge=lfs -text
|
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/llama/meta-llama/Llama-3.1-8B/3b69da34b19e6d21ec06.json
ADDED
|
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_entry_class": "SingleModelCacheEntry",
|
| 3 |
+
"_model_id": "meta-llama/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": 32,
|
| 22 |
+
"capacity_factor": null,
|
| 23 |
+
"checkpoint_id": "meta-llama/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": 4,
|
| 30 |
+
"max_batch_size": 32,
|
| 31 |
+
"max_context_length": 4096,
|
| 32 |
+
"max_topk": 256,
|
| 33 |
+
"n_active_tokens": 4096,
|
| 34 |
+
"neuronxcc_version": "2.21.33363.0+82129205",
|
| 35 |
+
"on_device_sampling": true,
|
| 36 |
+
"optimum_neuron_version": "0.4.4.dev2",
|
| 37 |
+
"output_logits": false,
|
| 38 |
+
"pp_degree": 1,
|
| 39 |
+
"sequence_length": 4096,
|
| 40 |
+
"speculation_length": 0,
|
| 41 |
+
"start_rank_id": 0,
|
| 42 |
+
"target": "trn2",
|
| 43 |
+
"torch_dtype": "bfloat16",
|
| 44 |
+
"tp_degree": 4
|
| 45 |
+
},
|
| 46 |
+
"num_attention_heads": 32,
|
| 47 |
+
"num_hidden_layers": 32,
|
| 48 |
+
"num_key_value_heads": 8,
|
| 49 |
+
"pretraining_tp": 1,
|
| 50 |
+
"rms_norm_eps": 1e-05,
|
| 51 |
+
"rope_scaling": {
|
| 52 |
+
"factor": 8.0,
|
| 53 |
+
"high_freq_factor": 4.0,
|
| 54 |
+
"low_freq_factor": 1.0,
|
| 55 |
+
"original_max_position_embeddings": 8192,
|
| 56 |
+
"rope_type": "llama3"
|
| 57 |
+
},
|
| 58 |
+
"rope_theta": 500000.0,
|
| 59 |
+
"tie_word_embeddings": false,
|
| 60 |
+
"use_cache": true,
|
| 61 |
+
"vocab_size": 128256
|
| 62 |
+
}
|
neuronxcc-2.21.33363.0+82129205/MODULE_3fba20491a2dd90d7fd7+ac10809c/compile_flags.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
["--target=trn2", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=2", "--logfile=/tmp/nxd_model/token_generation/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
|
neuronxcc-2.21.33363.0+82129205/MODULE_3fba20491a2dd90d7fd7+ac10809c/model.done
ADDED
|
File without changes
|
neuronxcc-2.21.33363.0+82129205/MODULE_3fba20491a2dd90d7fd7+ac10809c/model.hlo_module.pb
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1975439325968a460ccff1cbb522df9b9f77fbe55a96b9832c0d4483ae42af50
|
| 3 |
+
size 5577251
|
neuronxcc-2.21.33363.0+82129205/MODULE_3fba20491a2dd90d7fd7+ac10809c/model.neff
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0b51cbfe44a0c5344e5ce6e605fcbd9095c054bb3b38eba94bfa877d9851f370
|
| 3 |
+
size 15217664
|
neuronxcc-2.21.33363.0+82129205/MODULE_3fba20491a2dd90d7fd7+ac10809c/wrapped_neff.hlo
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:55f45f95ee2f0a00fd55d85c4b4f2bd9d4b953937330da9bcb90ef8ffe0d12e5
|
| 3 |
+
size 15368070
|
neuronxcc-2.21.33363.0+82129205/MODULE_ebdaf819ea1c1978f2f3+a32116a7/compile_flags.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
["--target=trn2", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=2", "--logfile=/tmp/nxd_model/context_encoding/_tp0_bk0/log-neuron-cc.txt"]
|
neuronxcc-2.21.33363.0+82129205/MODULE_ebdaf819ea1c1978f2f3+a32116a7/model.done
ADDED
|
File without changes
|
neuronxcc-2.21.33363.0+82129205/MODULE_ebdaf819ea1c1978f2f3+a32116a7/model.hlo_module.pb
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c537a1460d2ac7947f67f73eab497cc30d3e0ce3ae76cbc7fa240cb6c64c3bb0
|
| 3 |
+
size 2204489
|
neuronxcc-2.21.33363.0+82129205/MODULE_ebdaf819ea1c1978f2f3+a32116a7/model.neff
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:434336fee8bed168f97e7b07ca88b8650402bfe7b1c0ba92997aceadbf7aaade
|
| 3 |
+
size 3431424
|